Merge branch 'main' into next-v2

Reviewed-on: https://git.deuxfleurs.fr/Deuxfleurs/garage/pulls/1068

Cargo.lock

@@ -556,7 +556,7 @@ checksum = "b62ddb9cb1ec0a098ad4bbf9344d0713fa193ae1a80af55febcff2627b6a00c1"
 dependencies = [
  "getrandom 0.2.15",
  "instant",
- "rand",
+ "rand 0.8.5",
 ]
 
 [[package]]
@@ -838,6 +838,20 @@ version = "2.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "19d374276b40fb8bbdee95aef7c7fa6b5316ec764510eb64b8dd0e2ed0d7e7f5"
 
+[[package]]
+name = "crc-fast"
+version = "1.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "bf0ab672ed7761a44410115d2e0e5ccd86b987ba186a78a4148fb02bb302a774"
+dependencies = [
+ "cc",
+ "crc",
+ "digest",
+ "libc",
+ "rand 0.9.1",
+ "regex",
+]
+
 [[package]]
 name = "crc32c"
 version = "0.6.8"
@@ -896,7 +910,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "1bfb12502f3fc46cca1bb51ac28df9d618d813cdc3d2f25b9fe775a34af26bb3"
 dependencies = [
  "generic-array",
- "rand_core",
+ "rand_core 0.6.4",
  "typenum",
 ]
 
@@ -1211,7 +1225,7 @@ dependencies = [
  "bytes",
  "bytesize",
  "chrono",
- "crc32fast",
+ "crc-fast",
  "format_table",
  "futures",
  "garage_api_admin",
@@ -1249,6 +1263,7 @@ dependencies = [
  "timeago",
  "tokio",
  "tracing",
+ "tracing-journald",
  "tracing-subscriber",
  "utoipa",
 ]
@@ -1292,9 +1307,7 @@ dependencies = [
  "base64 0.21.7",
  "bytes",
  "chrono",
- "crc32c",
- "crc32fast",
- "crc64fast-nvme",
+ "crc-fast",
  "crypto-common",
  "err-derive",
  "futures",
@@ -1307,7 +1320,6 @@ dependencies = [
  "http-body-util",
  "hyper 1.6.0",
  "hyper-util",
- "idna 0.5.0",
  "md-5",
  "nom",
  "opentelemetry",
@@ -1353,9 +1365,7 @@ dependencies = [
  "base64 0.21.7",
  "bytes",
  "chrono",
- "crc32c",
- "crc32fast",
- "crc64fast-nvme",
+ "crc-fast",
  "err-derive",
  "form_urlencoded",
  "futures",
@@ -1407,7 +1417,7 @@ dependencies = [
  "garage_util",
  "hex",
  "opentelemetry",
- "rand",
+ "rand 0.8.5",
  "serde",
  "tokio",
  "tokio-util 0.7.14",
@@ -1448,7 +1458,7 @@ dependencies = [
  "hex",
  "http 1.3.1",
  "parse_duration",
- "rand",
+ "rand 0.8.5",
  "serde",
  "serde_bytes",
  "tokio",
@@ -1473,7 +1483,7 @@ dependencies = [
  "opentelemetry-contrib",
  "pin-project",
  "pretty_env_logger",
- "rand",
+ "rand 0.8.5",
  "rmp-serde",
  "serde",
  "tokio",
@@ -1503,7 +1513,7 @@ dependencies = [
  "nix",
  "opentelemetry",
  "pnet_datalink",
- "rand",
+ "rand 0.8.5",
  "reqwest",
  "schemars",
  "serde",
@@ -1527,7 +1537,7 @@ dependencies = [
  "hex",
  "hexdump",
  "opentelemetry",
- "rand",
+ "rand 0.8.5",
  "serde",
  "serde_bytes",
  "tokio",
@@ -1554,7 +1564,7 @@ dependencies = [
  "lazy_static",
  "mktemp",
  "opentelemetry",
- "rand",
+ "rand 0.8.5",
  "rmp-serde",
  "rustc_version",
  "serde",
@@ -2179,16 +2189,6 @@ version = "1.0.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b9e0384b61958566e926dc50660321d12159025e767c18e043daf26b70104c39"
 
-[[package]]
-name = "idna"
-version = "0.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "634d9b1461af396cad843f47fdba5597a4f9e6ddd4bfb6ff5d85028c25cb12f6"
-dependencies = [
- "unicode-bidi",
- "unicode-normalization",
-]
-
 [[package]]
 name = "idna"
 version = "1.0.3"
@@ -2868,7 +2868,7 @@ dependencies = [
  "lazy_static",
  "percent-encoding",
  "pin-project",
- "rand",
+ "rand 0.8.5",
  "thiserror 1.0.69",
  "tokio",
  "tokio-stream",
@@ -3010,7 +3010,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "346f04948ba92c43e8469c1ee6736c7563d71012b17d40745260fe106aac2166"
 dependencies = [
  "base64ct",
- "rand_core",
+ "rand_core 0.6.4",
  "subtle",
 ]
 
@@ -3359,8 +3359,18 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "34af8d1a0e25924bc5b7c43c079c942339d8f0a8b57c39049bef581b46327404"
 dependencies = [
  "libc",
- "rand_chacha",
- "rand_core",
+ "rand_chacha 0.3.1",
+ "rand_core 0.6.4",
+]
+
+[[package]]
+name = "rand"
+version = "0.9.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9fbfd9d094a40bf3ae768db9361049ace4c0e04a4fd6b359518bd7b73a73dd97"
+dependencies = [
+ "rand_chacha 0.9.0",
+ "rand_core 0.9.3",
 ]
 
 [[package]]
@@ -3370,7 +3380,17 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "e6c10a63a0fa32252be49d21e7709d4d4baf8d231c2dbce1eaa8141b9b127d88"
 dependencies = [
  "ppv-lite86",
- "rand_core",
+ "rand_core 0.6.4",
+]
+
+[[package]]
+name = "rand_chacha"
+version = "0.9.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d3022b5f1df60f26e1ffddd6c66e8aa15de382ae63b3a0c1bfc0e4d3e3f325cb"
+dependencies = [
+ "ppv-lite86",
+ "rand_core 0.9.3",
 ]
 
 [[package]]
@@ -3382,6 +3402,15 @@ dependencies = [
  "getrandom 0.2.15",
 ]
 
+[[package]]
+name = "rand_core"
+version = "0.9.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "99d9a13982dcf210057a8a78572b2217b667c3beacbf3a0d8b454f6f82837d38"
+dependencies = [
+ "getrandom 0.3.2",
+]
+
 [[package]]
 name = "redox_syscall"
 version = "0.2.16"
@@ -4262,21 +4291,6 @@ dependencies = [
  "zerovec",
 ]
 
-[[package]]
-name = "tinyvec"
-version = "1.9.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "09b3661f17e86524eccd4371ab0429194e0d7c008abb45f7a7495b1719463c71"
-dependencies = [
- "tinyvec_macros",
-]
-
-[[package]]
-name = "tinyvec_macros"
-version = "0.1.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1f3ccbac311fea05f86f61904b462b55fb3df8837a366dfc601a0161d0532f20"
-
 [[package]]
 name = "tokio"
 version = "1.44.1"
@@ -4465,7 +4479,7 @@ dependencies = [
  "indexmap 1.9.3",
  "pin-project",
  "pin-project-lite",
- "rand",
+ "rand 0.8.5",
  "slab",
  "tokio",
  "tokio-util 0.7.14",
@@ -4550,6 +4564,17 @@ dependencies = [
  "tracing",
 ]
 
+[[package]]
+name = "tracing-journald"
+version = "0.3.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "fc0b4143302cf1022dac868d521e36e8b27691f72c84b3311750d5188ebba657"
+dependencies = [
+ "libc",
+ "tracing-core",
+ "tracing-subscriber",
+]
+
 [[package]]
 name = "tracing-log"
 version = "0.2.0"
@@ -4597,27 +4622,12 @@ version = "0.1.7"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "2896d95c02a80c6d6a5d6e953d479f5ddf2dfdb6a244441010e373ac0fb88971"
 
-[[package]]
-name = "unicode-bidi"
-version = "0.3.18"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5c1cb5db39152898a79168971543b1cb5020dff7fe43c8dc468b0885f5e29df5"
-
 [[package]]
 name = "unicode-ident"
 version = "1.0.18"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "5a5f39404a5da50712a4c1eecf25e90dd62b613502b7e925fd4e4d19b5c96512"
 
-[[package]]
-name = "unicode-normalization"
-version = "0.1.24"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5033c97c4262335cded6d6fc3e5c18ab755e1a3dc96376350f3d8e9f009ad956"
-dependencies = [
- "tinyvec",
-]
-
 [[package]]
 name = "unicode-segmentation"
 version = "1.12.0"
@@ -4665,7 +4675,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "32f8b686cadd1473f4bd0117a5d28d36b1ade384ea9b5069a1c40aefed7fda60"
 dependencies = [
  "form_urlencoded",
- "idna 1.0.3",
+ "idna",
  "percent-encoding",
 ]
 
@@ -4717,7 +4727,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "79daa5ed5740825c40b389c5e50312b9c86df53fccd33f281df655642b43869d"
 dependencies = [
  "getrandom 0.2.15",
- "rand",
+ "rand 0.8.5",
 ]
 
 [[package]]

Cargo.toml

@@ -49,9 +49,7 @@ bytes = "1.0"
 bytesize = "1.1"
 cfg-if = "1.0"
 chrono = { version = "0.4", features = ["serde"] }
-crc32fast = "1.4"
-crc32c = "0.6"
-crc64fast-nvme = "1.2"
+crc-fast = "1.1"
 crypto-common = "0.1"
 err-derive = "0.3"
 gethostname = "0.4"
@@ -59,7 +57,6 @@ git-version = "0.3.4"
 hex = "0.4"
 hexdump = "0.1"
 hmac = "0.12"
-idna = "0.5"
 itertools = "0.12"
 ipnet = "2.9.0"
 lazy_static = "1.4"
@@ -86,6 +83,7 @@ pretty_env_logger = "0.5"
 structopt = { version = "0.3", default-features = false }
 syslog-tracing = "0.3"
 tracing = "0.1"
+tracing-journald = "0.3.1"
 tracing-subscriber = { version = "0.3", features = ["env-filter"] }
 
 heed = { version = "0.11", default-features = false, features = ["lmdb"] }

doc/api/garage-admin-v0.html

@@ -1,7 +1,7 @@
 <!DOCTYPE html>
 <html>
   <head>
-    <title>Garage adminstration API v0</title>
+    <title>Garage administration API v0</title>
     <!-- needed for adaptive design -->
     <meta charset="utf-8"/>
     <meta name="viewport" content="width=device-width, initial-scale=1">

doc/api/garage-admin-v1.html

@@ -1,7 +1,7 @@
 <!DOCTYPE html>
 <html>
   <head>
-    <title>Garage adminstration API v1</title>
+    <title>Garage administration API v1</title>
     <!-- needed for adaptive design -->
     <meta charset="utf-8"/>
     <meta name="viewport" content="width=device-width, initial-scale=1">

doc/api/garage-admin-v2.html

@@ -1,7 +1,7 @@
 <!DOCTYPE html>
 <html>
   <head>
-    <title>Garage adminstration API v2</title>
+    <title>Garage administration API v2</title>
     <!-- needed for adaptive design -->
     <meta charset="utf-8"/>
     <meta name="viewport" content="width=device-width, initial-scale=1">

doc/api/garage-admin-v2.json

@@ -3120,9 +3120,15 @@
           "blocksPurged",
           "objectsDeleted",
           "uploadsDeleted",
-          "versionsDeleted"
+          "versionsDeleted",
+          "blockRefsPurged"
         ],
         "properties": {
+          "blockRefsPurged": {
+            "type": "integer",
+            "format": "int64",
+            "minimum": 0
+          },
           "blocksPurged": {
             "type": "integer",
             "format": "int64",
@@ -3574,9 +3580,15 @@
                 "blocksPurged",
                 "objectsDeleted",
                 "uploadsDeleted",
-                "versionsDeleted"
+                "versionsDeleted",
+                "blockRefsPurged"
               ],
               "properties": {
+                "blockRefsPurged": {
+                  "type": "integer",
+                  "format": "int64",
+                  "minimum": 0
+                },
                 "blocksPurged": {
                   "type": "integer",
                   "format": "int64",
@@ -3973,6 +3985,12 @@
                 "$ref": "#/components/schemas/ScrubCommand"
               }
             }
+          },
+          {
+            "type": "string",
+            "enum": [
+              "aliases"
+            ]
           }
         ]
       },

doc/book/connect/apps/index.md

@@ -69,7 +69,7 @@ $CONFIG = array(
         'hostname' => '127.0.0.1', // Can also be a domain name, eg. garage.example.com
         'port' => 3900,            // Put your reverse proxy port or your S3 API port
         'use_ssl' => false,        // Set it to true if you have a TLS enabled reverse proxy
-        'region' => 'garage',      // Garage has only one region named "garage"
+        'region' => 'garage',      // Garage default region is named "garage", edit according to your cluster config
         'use_path_style' => true   // Garage supports only path style, must be set to true
     ],
 ],
@@ -135,7 +135,7 @@ bucket but doesn't also know the secret encryption key.
 *Click on the picture to zoom*
 
 Add a new external storage. Put what you want in "folder name" (eg. "shared"). Select "Amazon S3". Keep "Access Key" for the Authentication field.
-In Configuration, put your bucket name (eg. nextcloud), the host (eg. 127.0.0.1), the port (eg. 3900 or 443), the region (garage). Tick the SSL box if you have put an HTTPS proxy in front of garage. You must tick the "Path access" box and you must leave the "Legacy authentication (v2)" box empty. Put your Key ID (eg. GK...) and your Secret Key in the last two input boxes. Finally click on the tick symbol on the right of your screen.
+In Configuration, put your bucket name (eg. nextcloud), the host (eg. 127.0.0.1), the port (eg. 3900 or 443), the region ("garage" if you use the default, or the one your configured in your `garage.toml`). Tick the SSL box if you have put an HTTPS proxy in front of garage. You must tick the "Path access" box and you must leave the "Legacy authentication (v2)" box empty. Put your Key ID (eg. GK...) and your Secret Key in the last two input boxes. Finally click on the tick symbol on the right of your screen.
 
 Now go to your "Files" app and a new "linked folder" has appeared with the name you chose earlier (eg. "shared").
 
@@ -238,7 +238,7 @@ object_storage:
   # Put localhost only if you have a garage instance running on that node
   endpoint: 'http://localhost:3900' # or "garage.example.com" if you have TLS on port 443
 
-  # Garage supports only one region for now, named garage
+  # Garage default region is named "garage", edit according to your config
   region: 'garage'
 
   credentials:
@@ -441,7 +441,7 @@ media_storage_providers:
   store_synchronous: True  # do we want to wait that the file has been written before returning?
   config:
     bucket: matrix       # the name of our bucket, we chose matrix earlier
-    region_name: garage  # only "garage" is supported for the region field
+    region_name: garage  # "garage" by default, edit according to your cluster config
     endpoint_url: http://localhost:3900 # the path to the S3 endpoint
     access_key_id: "GKxxx" # your Key ID
     secret_access_key: "xxxx" # your Secret Key

doc/book/cookbook/ansible.md

@@ -8,18 +8,18 @@ have published Ansible roles.  We list them and compare them below.
 
 ## Comparison of Ansible roles
 
-| Feature                            | [ansible-role-garage](#zorun-ansible-role-garage) | [garage-docker-ansible-deploy](#moan0s-garage-docker-ansible-deploy) |
-|------------------------------------|---------------------------------------------|---------------------------------------------------------------|
-| **Runtime**                        | Systemd                                     | Docker                                                        |
-| **Target OS**                      | Any Linux                                   | Any Linux                                                     |
-| **Architecture**                   | amd64, arm64, i686                          | amd64, arm64                                                  |
-| **Additional software**            | None                                        | Traefik                                                       |
-| **Automatic node connection**      | ❌                                          | ✅                                                            |
-| **Layout management**              | ❌                                          | ✅                                                            |
-| **Manage buckets & keys**          | ❌                                          | ✅ (basic)                                                    |
-| **Allow custom Garage config**     | ✅                                          | ❌                                                            |
-| **Facilitate Garage upgrades**     | ✅                                          | ❌                                                            |
-| **Multiple instances on one host** | ✅                                          | ✅                                                            |
+| Feature                            | [ansible-role-garage](#zorun-ansible-role-garage) | [garage-docker-ansible-deploy](#moan0s-garage-docker-ansible-deploy) | [eddster ansible-role-garage](#eddster-ansible-role-garage) |
+|------------------------------------|---------------------------------------------|---------------------------------------------------------------|---------------------------------|
+| **Runtime**                        | Systemd                                     | Docker                                                        | Systemd                         |
+| **Target OS**                      | Any Linux                                   | Any Linux                                                     | Any Linux                       |
+| **Architecture**                   | amd64, arm64, i686                          | amd64, arm64                                                  | arm64, arm, 386, amd64                            |
+| **Additional software**            | None                                        | Traefik                                                       | Ngnix and Keepalived (optional) |
+| **Automatic node connection**      | ❌                                          | ✅                                                            | ✅                              |
+| **Layout management**              | ❌                                          | ✅                                                            | ✅                              |
+| **Manage buckets & keys**          | ❌                                          | ✅ (basic)                                                    | ✅                              |
+| **Allow custom Garage config**     | ✅                                          | ❌                                                            | ❌                              |
+| **Facilitate Garage upgrades**     | ✅                                          | ❌                                                            | ✅                              |
+| **Multiple instances on one host** | ✅                                          | ✅                                                            | ❌                              |
 
 
 ## zorun/ansible-role-garage
@@ -49,3 +49,15 @@ structured DNS names, etc).
 
 As a result, this role makes it easier to start with Garage on Ansible,
 but is less flexible.
+
+## eddster2309/ansible-role-garage
+
+[Source code](https://github.com/eddster2309/ansible-role-garage), [Ansible galaxy](https://galaxy.ansible.com/ui/standalone/roles/eddster2309/garage/)
+
+This role is a opinionated but customisable role using the official Garage
+static binaries and only requires Systemd. As such it should work on any
+Linux based host. It includes all the nesscary configuration to
+automatically setup a clustered Garage deployment. Most Garage
+configuration options are exposed through Ansible variables so while you
+can't provide a custom config you can get very close. It can optionally
+installed a HA nginx deployment with Keepalived.

doc/book/cookbook/kubernetes.md

@@ -26,6 +26,13 @@ Or deploy with custom values:
 helm install --create-namespace --namespace garage garage ./garage -f values.override.yaml
 ```
 
+If you want to manage the CustomRessourceDefinition used by garage for its `kubernetes_discovery` outside of the helm chart, add `garage.kubernetesSkipCrd: true` to your custom values and use the kustomization before deploying the helm chart:
+
+```bash
+kubectl apply -k ../k8s/crd
+helm install --create-namespace --namespace garage garage ./garage -f values.override.yaml
+```
+
 After deploying, cluster layout must be configured manually as described in [Creating a cluster layout](@/documentation/quick-start/_index.md#creating-a-cluster-layout). Use the following command to access garage CLI:
 
 ```bash
@@ -45,7 +52,7 @@ This is an example `values.overrride.yaml` for deploying in a microk8s cluster w
 ```yaml
 garage:
   # Use only 2 replicas per object
-  replicationMode: "2"
+  replicationFactor: 2
 
 # Start 4 instances (StatefulSets) of garage
 deployment:

doc/book/cookbook/systemd.md

@@ -28,6 +28,7 @@ StateDirectory=garage
 DynamicUser=true
 ProtectHome=true
 NoNewPrivileges=true
+LimitNOFILE=42000
 
 [Install]
 WantedBy=multi-user.target

doc/book/quick-start/_index.md

@@ -129,9 +129,9 @@ docker run \
   -d \
   --name garaged \
   -p 3900:3900 -p 3901:3901 -p 3902:3902 -p 3903:3903 \
-  -v /etc/garage.toml:/path/to/garage.toml \
-  -v /var/lib/garage/meta:/path/to/garage/meta \
-  -v /var/lib/garage/data:/path/to/garage/data \
+  -v /path/to/garage.toml:/etc/garage.toml \
+  -v /path/to/garage/meta:/var/lib/garage/meta \
+  -v /path/to/garage/data:/var/lib/garage/data \
   dxflrs/garage:v2.0.0
 ```
 
@@ -182,11 +182,12 @@ ID                 Hostname  Address         Tag                   Zone  Capacit
 ## Creating a cluster layout
 
 Creating a cluster layout for a Garage deployment means informing Garage
-of the disk space available on each node of the cluster
-as well as the zone (e.g. datacenter) each machine is located in.
+of the disk space available on each node of the cluster, `-c`,
+as well as the name of the zone (e.g. datacenter), `-z`, each machine is located in.
 
-For our test deployment, we are using only one node. The way in which we configure
-it does not matter, you can simply write:
+For our test deployment, we are have only one node with zone named `dc1` and a
+capacity of `1G`, though the capacity is ignored for a single node deployment
+and can be changed later when adding new nodes.
 
 ```bash
 garage layout assign -z dc1 -c 1G <node_id>

doc/book/reference-manual/configuration.md

@@ -46,6 +46,7 @@ bootstrap_peers = [
     "212fd62eeaca72c122b45a7f4fa0f55e012aa5e24ac384a72a3016413fa724ff@[fc00:F::1]:3901",
 ]
 
+allow_punycode = false
 
 [consul_discovery]
 api = "catalog"
@@ -93,29 +94,30 @@ The following gives details about each available configuration option.
 
 [Environment variables](#env_variables).
 
-Top-level configuration options:
+Top-level configuration options, in alphabetical order:
+[`allow_punycode`](#allow_punycode),
 [`allow_world_readable_secrets`](#allow_world_readable_secrets),
 [`block_ram_buffer_max`](#block_ram_buffer_max),
 [`block_size`](#block_size),
 [`bootstrap_peers`](#bootstrap_peers),
 [`compression_level`](#compression_level),
+[`consistency_mode`](#consistency_mode),
 [`data_dir`](#data_dir),
 [`data_fsync`](#data_fsync),
 [`db_engine`](#db_engine),
 [`disable_scrub`](#disable_scrub),
-[`use_local_tz`](#use_local_tz),
 [`lmdb_map_size`](#lmdb_map_size),
 [`metadata_auto_snapshot_interval`](#metadata_auto_snapshot_interval),
 [`metadata_dir`](#metadata_dir),
 [`metadata_fsync`](#metadata_fsync),
 [`metadata_snapshots_dir`](#metadata_snapshots_dir),
 [`replication_factor`](#replication_factor),
-[`consistency_mode`](#consistency_mode),
 [`rpc_bind_addr`](#rpc_bind_addr),
 [`rpc_bind_outgoing`](#rpc_bind_outgoing),
 [`rpc_public_addr`](#rpc_public_addr),
 [`rpc_public_addr_subnet`](#rpc_public_addr_subnet)
-[`rpc_secret`/`rpc_secret_file`](#rpc_secret).
+[`rpc_secret`/`rpc_secret_file`](#rpc_secret),
+[`use_local_tz`](#use_local_tz).
 
 The `[consul_discovery]` section:
 [`api`](#consul_api),
@@ -153,13 +155,17 @@ The `[admin]` section:
 
 ### Environment variables {#env_variables}
 
-The following configuration parameter must be specified as an environment
-variable, it does not exist in the configuration file:
+The following configuration parameters must be specified as environment variables,
+they do not exist in the configuration file:
 
 - `GARAGE_LOG_TO_SYSLOG` (since `v0.9.4`): set this to `1` or `true` to make the
   Garage daemon send its logs to `syslog` (using the libc `syslog` function)
   instead of printing to stderr.
 
+- `GARAGE_LOG_TO_JOURNALD` (since `v1.2.0`): set this to `1` or `true` to make the
+  Garage daemon send its logs to `journald` (using the native protocol of `systemd-journald`)
+  instead of printing to stderr.
+
 The following environment variables can be used to override the corresponding
 values in the configuration file:
 
@@ -171,7 +177,7 @@ values in the configuration file:
 
 ### Top-level configuration options
 
-#### `replication_factor` {#replication_factor}
+#### `replication_factor` (since `v1.0.0`) {#replication_factor}
 
 The replication factor can be any positive integer smaller or equal the node count in your cluster.
 The chosen replication factor has a big impact on the cluster's failure tolerancy and performance characteristics.
@@ -219,7 +225,7 @@ is in progress.  In theory, no data should be lost as rebalancing is a
 routine operation for Garage, although we cannot guarantee you that everything
  will go right in such an extreme scenario.
 
-#### `consistency_mode` {#consistency_mode}
+#### `consistency_mode` (since `v1.0.0`) {#consistency_mode}
 
 The consistency mode setting determines the read and write behaviour of your cluster.
 
@@ -606,7 +612,7 @@ be obtained by running `garage node id` and then included directly in the
 key will be returned by `garage node id` and you will have to add the IP
 yourself.
 
-### `allow_world_readable_secrets` or `GARAGE_ALLOW_WORLD_READABLE_SECRETS` (env) {#allow_world_readable_secrets}
+#### `allow_world_readable_secrets` or `GARAGE_ALLOW_WORLD_READABLE_SECRETS` (env) {#allow_world_readable_secrets}
 
 Garage checks the permissions of your secret files to make sure they're not
 world-readable. In some cases, the check might fail and consider your files as
@@ -618,6 +624,13 @@ permission verification.
 Alternatively, you can set the `GARAGE_ALLOW_WORLD_READABLE_SECRETS`
 environment variable to `true` to bypass the permissions check.
 
+#### `allow_punycode` {#allow_punycode}
+
+Allow creating buckets with names containing punycode. When used for buckets served
+as websites, this allows using almost any unicode character in the domain name.
+
+Default to `false`.
+
 ### The `[consul_discovery]` section
 
 Garage supports discovering other nodes of the cluster using Consul.  For this

doc/book/reference-manual/s3-compatibility.md

@@ -23,7 +23,6 @@ Feel free to open a PR to suggest fixes this table. Minio is missing because the
 - 2022-05-25 - Many Ceph S3 endpoints are not documented but implemented. Following a notification from the Ceph community, we added them.
 
 
-
 ## High-level features
 
 | Feature                      | Garage                           | [Openstack Swift](https://docs.openstack.org/swift/latest/s3_compat.html) | [Ceph Object Gateway](https://docs.ceph.com/en/latest/radosgw/s3/) | [Riak CS](https://docs.riak.com/riak/cs/2.1.1/references/apis/storage/s3/index.html) | [OpenIO](https://docs.openio.io/latest/source/arch-design/s3_compliancy.html) |
@@ -34,6 +33,7 @@ Feel free to open a PR to suggest fixes this table. Minio is missing because the
 | [URL vhost-style](https://docs.aws.amazon.com/AmazonS3/latest/userguide/VirtualHosting.html#virtual-hosted-style-access) URL (eg. `bucket.host.tld/key`) |  ✅ Implemented | ❌| ✅| ✅ | ✅ |
 | [Presigned URLs](https://docs.aws.amazon.com/AmazonS3/latest/userguide/ShareObjectPreSignedURL.html) |  ✅ Implemented | ❌|  ✅ | ✅ |  ✅(❓) |
 | [SSE-C encryption](https://docs.aws.amazon.com/AmazonS3/latest/userguide/ServerSideEncryptionCustomerKeys.html) |  ✅ Implemented | ❓ |  ✅ | ❌ |  ✅ |
+| [Bucket versioning](https://docs.aws.amazon.com/AmazonS3/latest/userguide/Versioning.html) | ❌ Missing | ✅ |  ✅ | ❌ | ✅ |
 
 *Note:* OpenIO does not says if it supports presigned URLs. Because it is part
 of signature v4 and they claim they support it without additional precisions,

doc/drafts/admin-api.md

@@ -15,7 +15,10 @@ when changes are introduced.
 
 The Garage administration API was introduced in version 0.7.2, and was
 changed several times.
-This document applies only to the Garage v2 API (starting with Garage v2.0.0).
+
+**THIS DOCUMENT IS DEPRECATED.** We now have an OpenAPI spec which is automatically generated
+from Garage's source code and is always up-to-date. See `doc/api/garage-admin-v2.html`.
+Text in this document is no longer kept in sync with the admin API's actual behavior.
 
 
 ## Access control

nix/compile.nix

@@ -74,6 +74,7 @@ let
       "metrics"
       "telemetry-otlp"
       "syslog"
+      "journald"
     ]));
 
   featuresStr = lib.concatStringsSep "," rootFeatures;

script/dev-cluster.sh

@@ -30,6 +30,12 @@ for count in $(seq 1 3); do
 CONF_PATH="/tmp/config.$count.toml"
 LABEL="\e[${FANCYCOLORS[$count]}[$count]\e[49m"
 
+if [ "$GARAGE_OLDVER" == "v08" ]; then
+	REPLICATION_MODE="replication_mode = \"3\""
+else
+	REPLICATION_MODE="replication_factor = 3"
+fi
+
 cat > $CONF_PATH <<EOF
 block_size = 1048576			# objects are split in blocks of maximum this number of bytes
 metadata_dir = "/tmp/garage-meta-$count"
@@ -38,7 +44,7 @@ data_dir = "/tmp/garage-data-$count"
 rpc_bind_addr = "0.0.0.0:$((3900+$count))"		# the port other Garage nodes will use to talk to this node
 rpc_public_addr = "127.0.0.1:$((3900+$count))"
 bootstrap_peers = []
-replication_mode = "3"
+$REPLICATION_MODE
 rpc_secret = "$NETWORK_SECRET"
 
 [s3_api]

script/helm/garage/Chart.yaml

@@ -1,24 +1,18 @@
 apiVersion: v2
 name: garage
 description: S3-compatible object store for small self-hosted geo-distributed deployments
-
-# A chart can be either an 'application' or a 'library' chart.
-#
-# Application charts are a collection of templates that can be packaged into versioned archives
-# to be deployed.
-#
-# Library charts provide useful utilities or functions for the chart developer. They're included as
-# a dependency of application charts to inject those utilities and functions into the rendering
-# pipeline. Library charts do not define any templates and therefore cannot be deployed.
 type: application
-
-# This is the chart version. This version number should be incremented each time you make changes
-# to the chart and its templates, including the app version.
-# Versions are expected to follow Semantic Versioning (https://semver.org/)
 version: 0.8.0
-
-# This is the version number of the application being deployed. This version number should be
-# incremented each time you make changes to the application. Versions are not expected to
-# follow Semantic Versioning. They should reflect the version the application is using.
-# It is recommended to use it with quotes.
 appVersion: "v2.0.0"
+home: https://garagehq.deuxfleurs.fr/
+icon: https://garagehq.deuxfleurs.fr/images/garage-logo.svg
+
+keywords:
+- geo-distributed
+- read-after-write-consistency
+- s3-compatible
+
+sources:
+- https://git.deuxfleurs.fr/Deuxfleurs/garage.git
+
+maintainers: []

script/helm/garage/README.md

@@ -1,9 +1,15 @@
 # garage
 
-![Version: 0.6.0](https://img.shields.io/badge/Version-0.6.0-informational?style=flat-square) ![Type: application](https://img.shields.io/badge/Type-application-informational?style=flat-square) ![AppVersion: v1.0.1](https://img.shields.io/badge/AppVersion-v1.0.1-informational?style=flat-square)
+![Version: 0.8.0](https://img.shields.io/badge/Version-0.8.0-informational?style=flat-square) ![Type: application](https://img.shields.io/badge/Type-application-informational?style=flat-square) ![AppVersion: v2.0.0](https://img.shields.io/badge/AppVersion-v2.0.0-informational?style=flat-square)
 
 S3-compatible object store for small self-hosted geo-distributed deployments
 
+**Homepage:** <https://garagehq.deuxfleurs.fr/>
+
+## Source Code
+
+* <https://git.deuxfleurs.fr/Deuxfleurs/garage.git>
+
 ## Values
 
 | Key | Type | Default | Description |
@@ -16,14 +22,16 @@ S3-compatible object store for small self-hosted geo-distributed deployments
 | extraVolumeMounts | object | `{}` |  |
 | extraVolumes | object | `{}` |  |
 | fullnameOverride | string | `""` |  |
-| garage.blockSize | string | `"1048576"` | Defaults is 1MB An increase can result in better performance in certain scenarios https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#block-size |
+| garage.blockSize | string | `"1048576"` | Defaults is 1MB An increase can result in better performance in certain scenarios https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#block_size |
 | garage.bootstrapPeers | list | `[]` | This is not required if you use the integrated kubernetes discovery |
-| garage.compressionLevel | string | `"1"` | zstd compression level of stored blocks https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#compression-level |
-| garage.dbEngine | string | `"lmdb"` | Can be changed for better performance on certain systems https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#db-engine-since-v0-8-0 |
+| garage.compressionLevel | string | `"1"` | zstd compression level of stored blocks https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#compression_level |
+| garage.dbEngine | string | `"lmdb"` | Can be changed for better performance on certain systems https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#db_engine |
 | garage.existingConfigMap | string | `""` | if not empty string, allow using an existing ConfigMap for the garage.toml, if set, ignores garage.toml |
 | garage.garageTomlString | string | `""` | String Template for the garage configuration if set, ignores above values. Values can be templated, see https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/ |
 | garage.kubernetesSkipCrd | bool | `false` | Set to true if you want to use k8s discovery but install the CRDs manually outside of the helm chart, for example if you operate at namespace level without cluster ressources |
-| garage.replicationMode | string | `"3"` | Default to 3 replicas, see the replication_mode section at https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#replication-mode |
+| garage.replicationFactor | string | `"3"` | Default to 3 replicas, see the replication_factor section at https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#replication_factor |
+| garage.consistencyMode | string | `"consistent"` | Default to read-after-write consistency, see the consistency_mode section at https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#consistency_mode |
+| garage.metadataAutoSnapshotInterval | string | `""` | If this value is set, Garage will automatically take a snapshot of the metadata DB file at a regular interval and save it in the metadata directory. https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#metadata_auto_snapshot_interval |
 | garage.rpcBindAddr | string | `"[::]:3901"` |  |
 | garage.rpcSecret | string | `""` | If not given, a random secret will be generated and stored in a Secret object |
 | garage.s3.api.region | string | `"garage"` |  |
@@ -49,6 +57,7 @@ S3-compatible object store for small self-hosted geo-distributed deployments
 | initImage.pullPolicy | string | `"IfNotPresent"` |  |
 | initImage.repository | string | `"busybox"` |  |
 | initImage.tag | string | `"stable"` |  |
+| livenessProbe | object | `{}` | Specifies a livenessProbe |
 | monitoring.metrics.enabled | bool | `false` | If true, a service for monitoring is created with a prometheus.io/scrape annotation |
 | monitoring.metrics.serviceMonitor.enabled | bool | `false` | If true, a ServiceMonitor CRD is created for a prometheus operator https://github.com/coreos/prometheus-operator |
 | monitoring.metrics.serviceMonitor.interval | string | `"15s"` |  |
@@ -71,6 +80,7 @@ S3-compatible object store for small self-hosted geo-distributed deployments
 | podSecurityContext.runAsGroup | int | `1000` |  |
 | podSecurityContext.runAsNonRoot | bool | `true` |  |
 | podSecurityContext.runAsUser | int | `1000` |  |
+| readinessProbe | object | `{}` | Specifies a readinessProbe |
 | resources | object | `{}` |  |
 | securityContext.capabilities | object | `{"drop":["ALL"]}` | The default security context is heavily restricted, feel free to tune it to your requirements |
 | securityContext.readOnlyRootFilesystem | bool | `true` |  |

script/helm/garage/templates/configmap.yaml

@@ -15,10 +15,15 @@ data:
 
     block_size = {{ .Values.garage.blockSize }}
 
-    replication_mode = "{{ .Values.garage.replicationMode }}"
+    replication_factor = {{ .Values.garage.replicationFactor }}
+    consistency_mode = "{{ .Values.garage.consistencyMode }}"
 
     compression_level = {{ .Values.garage.compressionLevel }}
 
+    {{- if .Values.garage.metadataAutoSnapshotInterval }}
+    metadata_auto_snapshot_interval = {{ .Values.garage.metadataAutoSnapshotInterval | quote }}
+    {{- end }}
+
     rpc_bind_addr = "{{ .Values.garage.rpcBindAddr }}"
     # rpc_secret will be populated by the init container from a k8s secret object
     rpc_secret = "__RPC_SECRET_REPLACE__"

script/helm/garage/templates/workload.yaml

@@ -78,15 +78,14 @@ spec:
           {{- with .Values.extraVolumeMounts }}
             {{- toYaml . | nindent 12 }}
           {{- end }}
-          # TODO
-          # livenessProbe:
-          #   httpGet:
-          #     path: /
-          #     port: 3900
-          # readinessProbe:
-          #   httpGet:
-          #     path: /
-          #     port: 3900
+          {{- with .Values.livenessProbe }}
+          livenessProbe:
+            {{- toYaml . | nindent 12 }}
+          {{- end }}
+          {{- with .Values.readinessProbe }}
+          readinessProbe:
+            {{- toYaml . | nindent 12 }}
+          {{- end }}
           resources:
             {{- toYaml .Values.resources | nindent 12 }}
       volumes:

script/helm/garage/values.yaml

@@ -5,22 +5,30 @@
 # Garage configuration. These values go to garage.toml
 garage:
   # -- Can be changed for better performance on certain systems
-  # https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#db-engine-since-v0-8-0
+  # https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#db_engine
   dbEngine: "lmdb"
 
   # -- Defaults is 1MB
   # An increase can result in better performance in certain scenarios
-  # https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#block-size
+  # https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#block_size
   blockSize: "1048576"
 
-  # -- Default to 3 replicas, see the replication_mode section at
-  # https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#replication-mode
-  replicationMode: "3"
+  # -- Default to 3 replicas, see the replication_factor section at
+  # https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#replication_factor
+  replicationFactor: "3"
+
+  # -- By default, enable read-after-write consistency guarantees, see the consistency_mode section at
+  # https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#consistency_mode
+  consistencyMode: "consistent"
 
   # -- zstd compression level of stored blocks
-  # https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#compression-level
+  # https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#compression_level
   compressionLevel: "1"
 
+  # -- If this value is set, Garage will automatically take a snapshot of the metadata DB file at a regular interval and save it in the metadata directory.
+  # https://garagehq.deuxfleurs.fr/documentation/reference-manual/configuration/#metadata_auto_snapshot_interval
+  metadataAutoSnapshotInterval: ""
+
   rpcBindAddr: "[::]:3901"
   # -- If not given, a random secret will be generated and stored in a Secret object
   rpcSecret: ""
@@ -191,6 +199,21 @@ resources: {}
   #   cpu: 100m
   #   memory: 512Mi
 
+# -- Specifies a livenessProbe
+livenessProbe: {}
+  #httpGet:
+  #  path: /health
+  #  port: 3903
+  #initialDelaySeconds: 5
+  #periodSeconds: 30
+# -- Specifies a readinessProbe
+readinessProbe: {}
+  #httpGet:
+  #  path: /health
+  #  port: 3903
+  #initialDelaySeconds: 5
+  #periodSeconds: 30
+
 nodeSelector: {}
 
 tolerations: []

script/jepsen.garage/src/jepsen/garage/daemon.clj

@@ -43,7 +43,7 @@
              "rpc_bind_addr = \"0.0.0.0:3901\"\n"
              "rpc_public_addr = \"" node ":3901\"\n"
              "db_engine = \"lmdb\"\n"
-             "replication_mode = \"3\"\n"
+             "replication_factor = 3\n"
              "data_dir = \"" data-dir "\"\n"
              "metadata_dir = \"" meta-dir "\"\n"
              "[s3_api]\n"

script/k8s/config.yaml

@@ -8,7 +8,7 @@ data:
     metadata_dir = "/tmp/meta"
     data_dir = "/tmp/data"
 
-    replication_mode = "3"
+    replication_factor = 3
 
     rpc_bind_addr = "[::]:3901"
     rpc_secret = "1799bccfd7411eddcf9ebd316bc1f5287ad12a68094e1c6ac6abde7e6feae1ec"

script/k8s/crd/garagenodes.deuxfleurs.fr.yaml

@@ -0,0 +1,43 @@
+apiVersion: apiextensions.k8s.io/v1
+kind: CustomResourceDefinition
+metadata:
+  name: garagenodes.deuxfleurs.fr
+spec:
+  conversion:
+    strategy: None
+  group: deuxfleurs.fr
+  names:
+    kind: GarageNode
+    listKind: GarageNodeList
+    plural: garagenodes
+    singular: garagenode
+  scope: Namespaced
+  versions:
+  - name: v1
+    schema:
+      openAPIV3Schema:
+        description: Auto-generated derived type for Node via `CustomResource`
+        properties:
+          spec:
+            properties:
+              address:
+                format: ip
+                type: string
+              hostname:
+                type: string
+              port:
+                format: uint16
+                minimum: 0
+                type: integer
+            required:
+            - address
+            - hostname
+            - port
+            type: object
+        required:
+        - spec
+        title: GarageNode
+        type: object
+    served: true
+    storage: true
+    subresources: {}

script/k8s/crd/kustomization.yaml

@@ -0,0 +1,5 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+
+resources:
+- garagenodes.deuxfleurs.fr.yaml

script/test-smoke.sh

@@ -119,8 +119,12 @@ if [ -z "$SKIP_BOTO3" ]; then
   AWS_ENDPOINT_URL=https://localhost:4443 python <<EOF
 import boto3
 client = boto3.client('s3', verify=False)
+print("Put&delete hello world object")
 client.put_object(Body=b'hello world', Bucket='eprouvette', Key='test.s3.txt')
 client.delete_object(Bucket='eprouvette', Key='test.s3.txt')
+print("Put&delete big object")
+client.upload_file("/tmp/garage.3.rnd", 'eprouvette', 'garage.3.rnd')
+client.delete_object(Bucket='eprouvette', Key='garage.3.rnd')
 print("OK!")
 EOF
 fi

shell.nix

@@ -37,6 +37,7 @@ in
     ];
     shellHook = ''
       function to_s3 {
+        AWS_REQUEST_CHECKSUM_CALCULATION=WHEN_REQUIRED AWS_RESPONSE_CHECKSUM_VALIDATION=WHEN_REQUIRED \
         aws \
             --endpoint-url https://garage.deuxfleurs.fr \
             --region garage \
@@ -93,6 +94,7 @@ in
 
         nix-build nix/build_index.nix
 
+        AWS_REQUEST_CHECKSUM_CALCULATION=WHEN_REQUIRED AWS_RESPONSE_CHECKSUM_VALIDATION=WHEN_REQUIRED \
         aws \
             --endpoint-url https://garage.deuxfleurs.fr \
             --region garage \
@@ -100,6 +102,7 @@ in
             result/share/_releases.json \
             s3://garagehq.deuxfleurs.fr/
 
+        AWS_REQUEST_CHECKSUM_CALCULATION=WHEN_REQUIRED AWS_RESPONSE_CHECKSUM_VALIDATION=WHEN_REQUIRED \
         aws \
             --endpoint-url https://garage.deuxfleurs.fr \
             --region garage \

src/api/admin/api.rs

@@ -1114,6 +1114,7 @@ pub enum RepairType {
 	BlockRc,
 	Rebalance,
 	Scrub(ScrubCommand),
+	Aliases,
 }
 
 #[derive(Debug, Clone, Serialize, Deserialize, ToSchema)]
@@ -1308,4 +1309,5 @@ pub struct LocalPurgeBlocksResponse {
 	pub objects_deleted: u64,
 	pub uploads_deleted: u64,
 	pub versions_deleted: u64,
+	pub block_refs_purged: u64,
 }

src/api/admin/block.rs

@@ -151,6 +151,7 @@ impl RequestHandler for LocalPurgeBlocksRequest {
 		let mut obj_dels = 0;
 		let mut mpu_dels = 0;
 		let mut ver_dels = 0;
+		let mut br_dels = 0;
 
 		for hash in self.0.iter() {
 			let hash = hex::decode(hash).ok_or_bad_request("invalid hash")?;
@@ -176,11 +177,18 @@ impl RequestHandler for LocalPurgeBlocksRequest {
 						ver_dels += 1;
 					}
 				}
+				if !br.deleted.get() {
+					let mut br = br;
+					br.deleted.set();
+					garage.block_ref_table.insert(&br).await?;
+					br_dels += 1;
+				}
 			}
 		}
 
 		Ok(LocalPurgeBlocksResponse {
 			blocks_purged: self.0.len() as u64,
+			block_refs_purged: br_dels,
 			versions_deleted: ver_dels,
 			objects_deleted: obj_dels,
 			uploads_deleted: mpu_dels,

src/api/admin/bucket.rs

@@ -159,7 +159,7 @@ impl RequestHandler for CreateBucketRequest {
 		let helper = garage.locked_helper().await;
 
 		if let Some(ga) = &self.global_alias {
-			if !is_valid_bucket_name(ga) {
+			if !is_valid_bucket_name(ga, garage.config.allow_punycode) {
 				return Err(Error::bad_request(format!(
 					"{}: {}",
 					ga, INVALID_BUCKET_NAME_MESSAGE
@@ -174,7 +174,7 @@ impl RequestHandler for CreateBucketRequest {
 		}
 
 		if let Some(la) = &self.local_alias {
-			if !is_valid_bucket_name(&la.alias) {
+			if !is_valid_bucket_name(&la.alias, garage.config.allow_punycode) {
 				return Err(Error::bad_request(format!(
 					"{}: {}",
 					la.alias, INVALID_BUCKET_NAME_MESSAGE
@@ -255,7 +255,7 @@ impl RequestHandler for DeleteBucketRequest {
 		for ((key_id, alias), _, active) in state.local_aliases.items().iter() {
 			if *active {
 				helper
-					.unset_local_bucket_alias(bucket.id, key_id, alias)
+					.purge_local_bucket_alias(bucket.id, key_id, alias)
 					.await?;
 			}
 		}

src/api/admin/repair.rs

@@ -89,6 +89,10 @@ impl RequestHandler for LocalLaunchRepairOperationRequest {
 					garage.block_manager.clone(),
 				));
 			}
+			RepairType::Aliases => {
+				info!("Repairing bucket aliases (foreground)");
+				garage.locked_helper().await.repair_aliases().await?;
+			}
 		}
 		Ok(LocalLaunchRepairOperationResponse)
 	}

src/api/common/Cargo.toml

@@ -21,15 +21,12 @@ garage_util.workspace = true
 base64.workspace = true
 bytes.workspace = true
 chrono.workspace = true
-crc32fast.workspace = true
-crc32c.workspace = true
-crc64fast-nvme.workspace = true
+crc-fast.workspace = true
 crypto-common.workspace = true
 err-derive.workspace = true
 hex.workspace = true
 hmac.workspace = true
 md-5.workspace = true
-idna.workspace = true
 tracing.workspace = true
 nom.workspace = true
 pin-project.workspace = true

src/api/common/helpers.rs

@@ -8,7 +8,6 @@ use hyper::{
 	body::{Body, Bytes},
 	Request, Response,
 };
-use idna::domain_to_unicode;
 use serde::{Deserialize, Serialize};
 
 use garage_model::bucket_table::BucketParams;
@@ -97,7 +96,7 @@ pub fn authority_to_host(authority: &str) -> Result<String, Error> {
 			authority
 		))),
 	};
-	authority.map(|h| domain_to_unicode(h).0)
+	authority.map(|h| h.to_ascii_lowercase())
 }
 
 /// Extract the bucket name and the key name from an HTTP path and possibly a bucket provided in

src/api/common/signature/checksum.rs

@@ -1,10 +1,7 @@
-use std::convert::{TryFrom, TryInto};
-use std::hash::Hasher;
+use std::convert::TryInto;
 
 use base64::prelude::*;
-use crc32c::Crc32cHasher as Crc32c;
-use crc32fast::Hasher as Crc32;
-use crc64fast_nvme::Digest as Crc64Nvme;
+use crc_fast::{CrcAlgorithm, Digest as CrcDigest};
 use md5::{Digest, Md5};
 use sha1::Sha1;
 use sha2::Sha256;
@@ -22,6 +19,7 @@ pub const CONTENT_MD5: HeaderName = HeaderName::from_static("content-md5");
 pub const X_AMZ_CHECKSUM_ALGORITHM: HeaderName =
 	HeaderName::from_static("x-amz-checksum-algorithm");
 pub const X_AMZ_CHECKSUM_MODE: HeaderName = HeaderName::from_static("x-amz-checksum-mode");
+pub const X_AMZ_CHECKSUM_TYPE: HeaderName = HeaderName::from_static("x-amz-checksum-type");
 pub const X_AMZ_CHECKSUM_CRC32: HeaderName = HeaderName::from_static("x-amz-checksum-crc32");
 pub const X_AMZ_CHECKSUM_CRC32C: HeaderName = HeaderName::from_static("x-amz-checksum-crc32c");
 pub const X_AMZ_CHECKSUM_CRC64NVME: HeaderName =
@@ -29,6 +27,10 @@ pub const X_AMZ_CHECKSUM_CRC64NVME: HeaderName =
 pub const X_AMZ_CHECKSUM_SHA1: HeaderName = HeaderName::from_static("x-amz-checksum-sha1");
 pub const X_AMZ_CHECKSUM_SHA256: HeaderName = HeaderName::from_static("x-amz-checksum-sha256");
 
+// Values for x-amz-checksum-type
+pub const COMPOSITE: &str = "COMPOSITE";
+pub const FULL_OBJECT: &str = "FULL_OBJECT";
+
 pub type Crc32Checksum = [u8; 4];
 pub type Crc32cChecksum = [u8; 4];
 pub type Crc64NvmeChecksum = [u8; 8];
@@ -36,6 +38,21 @@ pub type Md5Checksum = [u8; 16];
 pub type Sha1Checksum = [u8; 20];
 pub type Sha256Checksum = [u8; 32];
 
+// -- MAP OF CRC ALGORITHMS :
+// CRC32        -> CrcAlgorithm::Crc32IsoHdlc
+// CRC32C       -> CrcAlgorithm::Crc32Iscsi
+// CRC64NVME    -> CrcAlgorithm::Crc64Nvme
+
+pub fn new_crc32() -> CrcDigest {
+	CrcDigest::new(CrcAlgorithm::Crc32IsoHdlc)
+}
+pub fn new_crc32c() -> CrcDigest {
+	CrcDigest::new(CrcAlgorithm::Crc32Iscsi)
+}
+pub fn new_crc64nvme() -> CrcDigest {
+	CrcDigest::new(CrcAlgorithm::Crc64Nvme)
+}
+
 #[derive(Debug, Default, Clone)]
 pub struct ExpectedChecksums {
 	// base64-encoded md5 (content-md5 header)
@@ -47,9 +64,9 @@ pub struct ExpectedChecksums {
 }
 
 pub struct Checksummer {
-	pub crc32: Option<Crc32>,
-	pub crc32c: Option<Crc32c>,
-	pub crc64nvme: Option<Crc64Nvme>,
+	pub crc32: Option<CrcDigest>,
+	pub crc32c: Option<CrcDigest>,
+	pub crc64nvme: Option<CrcDigest>,
 	pub md5: Option<Md5>,
 	pub sha1: Option<Sha1>,
 	pub sha256: Option<Sha256>,
@@ -98,13 +115,13 @@ impl Checksummer {
 			self.sha256 = Some(Sha256::new());
 		}
 		if matches!(&expected.extra, Some(ChecksumValue::Crc32(_))) {
-			self.crc32 = Some(Crc32::new());
+			self.crc32 = Some(new_crc32());
 		}
 		if matches!(&expected.extra, Some(ChecksumValue::Crc32c(_))) {
-			self.crc32c = Some(Crc32c::default());
+			self.crc32c = Some(new_crc32c());
 		}
 		if matches!(&expected.extra, Some(ChecksumValue::Crc64Nvme(_))) {
-			self.crc64nvme = Some(Crc64Nvme::default());
+			self.crc64nvme = Some(new_crc64nvme());
 		}
 		if matches!(&expected.extra, Some(ChecksumValue::Sha1(_))) {
 			self.sha1 = Some(Sha1::new());
@@ -114,13 +131,13 @@ impl Checksummer {
 	pub fn add(mut self, algo: Option<ChecksumAlgorithm>) -> Self {
 		match algo {
 			Some(ChecksumAlgorithm::Crc32) => {
-				self.crc32 = Some(Crc32::new());
+				self.crc32 = Some(new_crc32());
 			}
 			Some(ChecksumAlgorithm::Crc32c) => {
-				self.crc32c = Some(Crc32c::default());
+				self.crc32c = Some(new_crc32c());
 			}
 			Some(ChecksumAlgorithm::Crc64Nvme) => {
-				self.crc64nvme = Some(Crc64Nvme::default());
+				self.crc64nvme = Some(new_crc64nvme());
 			}
 			Some(ChecksumAlgorithm::Sha1) => {
 				self.sha1 = Some(Sha1::new());
@@ -138,10 +155,10 @@ impl Checksummer {
 			crc32.update(bytes);
 		}
 		if let Some(crc32c) = &mut self.crc32c {
-			crc32c.write(bytes);
+			crc32c.update(bytes);
 		}
 		if let Some(crc64nvme) = &mut self.crc64nvme {
-			crc64nvme.write(bytes);
+			crc64nvme.update(bytes);
 		}
 		if let Some(md5) = &mut self.md5 {
 			md5.update(bytes);
@@ -156,11 +173,9 @@ impl Checksummer {
 
 	pub fn finalize(self) -> Checksums {
 		Checksums {
-			crc32: self.crc32.map(|x| u32::to_be_bytes(x.finalize())),
-			crc32c: self
-				.crc32c
-				.map(|x| u32::to_be_bytes(u32::try_from(x.finish()).unwrap())),
-			crc64nvme: self.crc64nvme.map(|x| u64::to_be_bytes(x.sum64())),
+			crc32: self.crc32.map(|x| u32::to_be_bytes(x.finalize() as u32)),
+			crc32c: self.crc32c.map(|x| u32::to_be_bytes(x.finalize() as u32)),
+			crc64nvme: self.crc64nvme.map(|x| u64::to_be_bytes(x.finalize())),
 			md5: self.md5.map(|x| x.finalize()[..].try_into().unwrap()),
 			sha1: self.sha1.map(|x| x.finalize()[..].try_into().unwrap()),
 			sha256: self.sha256.map(|x| x.finalize()[..].try_into().unwrap()),
@@ -192,10 +207,11 @@ impl Checksums {
 		}
 		if let Some(extra) = expected.extra {
 			let algo = extra.algorithm();
-			if self.extract(Some(algo)) != Some(extra) {
+			let calculated = self.extract(Some(algo));
+			if calculated != Some(extra) {
 				return Err(Error::InvalidDigest(format!(
-					"Failed to validate checksum for algorithm {:?}",
-					algo
+					"Failed to validate checksum for algorithm {:?}: calculated {:?}, expected {:?}",
+					algo, calculated, extra
 				)));
 			}
 		}

src/api/s3/Cargo.toml

@@ -27,9 +27,7 @@ async-compression.workspace = true
 base64.workspace = true
 bytes.workspace = true
 chrono.workspace = true
-crc32fast.workspace = true
-crc32c.workspace = true
-crc64fast-nvme.workspace = true
+crc-fast.workspace = true
 err-derive.workspace = true
 hex.workspace = true
 hmac.workspace = true

src/api/s3/api_server.rs

@@ -223,6 +223,7 @@ impl ApiHandler for S3ApiServer {
 			Endpoint::DeleteBucket {} => handle_delete_bucket(ctx).await,
 			Endpoint::GetBucketLocation {} => handle_get_bucket_location(ctx),
 			Endpoint::GetBucketVersioning {} => handle_get_bucket_versioning(),
+			Endpoint::GetBucketAcl {} => handle_get_bucket_acl(ctx),
 			Endpoint::ListObjects {
 				delimiter,
 				encoding_type,

src/api/s3/bucket.rs

@@ -5,7 +5,7 @@ use hyper::{Request, Response, StatusCode};
 use garage_model::bucket_alias_table::*;
 use garage_model::bucket_table::Bucket;
 use garage_model::garage::Garage;
-use garage_model::key_table::Key;
+use garage_model::key_table::{Key, KeyParams};
 use garage_model::permission::BucketKeyPerm;
 use garage_table::util::*;
 use garage_util::crdt::*;
@@ -44,6 +44,55 @@ pub fn handle_get_bucket_versioning() -> Result<Response<ResBody>, Error> {
 		.body(string_body(xml))?)
 }
 
+pub fn handle_get_bucket_acl(ctx: ReqCtx) -> Result<Response<ResBody>, Error> {
+	let ReqCtx {
+		bucket_id, api_key, ..
+	} = ctx;
+	let key_p = api_key.params().ok_or_internal_error(
+		"Key should not be in deleted state at this point (in handle_get_bucket_acl)",
+	)?;
+
+	let mut grants: Vec<s3_xml::Grant> = vec![];
+	let kp = api_key.bucket_permissions(&bucket_id);
+
+	if kp.allow_owner {
+		grants.push(s3_xml::Grant {
+			grantee: create_grantee(&key_p, &api_key),
+			permission: s3_xml::Value("FULL_CONTROL".to_string()),
+		});
+	} else {
+		if kp.allow_read {
+			grants.push(s3_xml::Grant {
+				grantee: create_grantee(&key_p, &api_key),
+				permission: s3_xml::Value("READ".to_string()),
+			});
+			grants.push(s3_xml::Grant {
+				grantee: create_grantee(&key_p, &api_key),
+				permission: s3_xml::Value("READ_ACP".to_string()),
+			});
+		}
+		if kp.allow_write {
+			grants.push(s3_xml::Grant {
+				grantee: create_grantee(&key_p, &api_key),
+				permission: s3_xml::Value("WRITE".to_string()),
+			});
+		}
+	}
+
+	let access_control_policy = s3_xml::AccessControlPolicy {
+		xmlns: (),
+		owner: None,
+		acl: s3_xml::AccessControlList { entries: grants },
+	};
+
+	let xml = s3_xml::to_xml_with_header(&access_control_policy)?;
+	trace!("xml: {}", xml);
+
+	Ok(Response::builder()
+		.header("Content-Type", "application/xml")
+		.body(string_body(xml))?)
+}
+
 pub async fn handle_list_buckets(
 	garage: &Garage,
 	api_key: &Key,
@@ -167,7 +216,7 @@ pub async fn handle_create_bucket(
 		}
 
 		// Create the bucket!
-		if !is_valid_bucket_name(&bucket_name) {
+		if !is_valid_bucket_name(&bucket_name, garage.config.allow_punycode) {
 			return Err(Error::bad_request(format!(
 				"{}: {}",
 				bucket_name, INVALID_BUCKET_NAME_MESSAGE
@@ -236,11 +285,11 @@ pub async fn handle_delete_bucket(ctx: ReqCtx) -> Result<Response<ResBody>, Erro
 		// 1. delete bucket alias
 		if is_local_alias {
 			helper
-				.unset_local_bucket_alias(*bucket_id, &api_key.key_id, bucket_name)
+				.purge_local_bucket_alias(*bucket_id, &api_key.key_id, bucket_name)
 				.await?;
 		} else {
 			helper
-				.unset_global_bucket_alias(*bucket_id, bucket_name)
+				.purge_global_bucket_alias(*bucket_id, bucket_name)
 				.await?;
 		}
 
@@ -306,6 +355,15 @@ fn parse_create_bucket_xml(xml_bytes: &[u8]) -> Option<Option<String>> {
 	Some(ret)
 }
 
+fn create_grantee(key_params: &KeyParams, api_key: &Key) -> s3_xml::Grantee {
+	s3_xml::Grantee {
+		xmlns_xsi: (),
+		typ: "CanonicalUser".to_string(),
+		display_name: Some(s3_xml::Value(key_params.name.get().to_string())),
+		id: Some(s3_xml::Value(api_key.key_id.to_string())),
+	}
+}
+
 #[cfg(test)]
 mod tests {
 	use super::*;

src/api/s3/copy.rs

@@ -26,9 +26,10 @@ use garage_api_common::signature::checksum::*;
 use crate::api_server::{ReqBody, ResBody};
 use crate::encryption::{EncryptionParams, OekDerivationInfo};
 use crate::error::*;
-use crate::get::{full_object_byte_stream, PreconditionHeaders};
+use crate::get::{check_version_not_deleted, full_object_byte_stream, PreconditionHeaders};
 use crate::multipart;
 use crate::put::{extract_metadata_headers, save_stream, ChecksumMode, SaveStreamResult};
+use crate::website::X_AMZ_WEBSITE_REDIRECT_LOCATION;
 use crate::xml::{self as s3_xml, xmlns_tag};
 
 pub const X_AMZ_COPY_SOURCE_IF_MATCH: HeaderName =
@@ -78,8 +79,24 @@ pub async fn handle_copy(
 		},
 	)?;
 
+	let was_multipart = source_version_meta.etag.contains('-') // HACK
+        || source_object_meta_inner.checksum_type == Some(ChecksumType::Composite);
+
 	// Extract source checksum info before source_object_meta_inner is consumed
 	let source_checksum = source_object_meta_inner.checksum;
+	let source_checksum_type = match (source_object_meta_inner.checksum_type, source_checksum) {
+		(Some(ct), _) => Some(ct),
+		(None, Some(_)) => {
+			// Migrated object from garage v1.x or older
+			// determine checksum type depending if this is a multipart upload or not
+			if was_multipart {
+				Some(ChecksumType::Composite)
+			} else {
+				Some(ChecksumType::FullObject)
+			}
+		}
+		(None, None) => None,
+	};
 	let source_checksum_algorithm = source_checksum.map(|x| x.algorithm());
 
 	// If source object has a checksum, the destination object must as well.
@@ -88,15 +105,26 @@ pub async fn handle_copy(
 	let checksum_algorithm = checksum_algorithm.or(source_checksum_algorithm);
 
 	// Determine metadata of destination object
-	let was_multipart = source_version_meta.etag.contains('-');
 	let dest_object_meta = ObjectVersionMetaInner {
 		headers: match req.headers().get("x-amz-metadata-directive") {
 			Some(v) if v == hyper::header::HeaderValue::from_static("REPLACE") => {
 				extract_metadata_headers(req.headers())?
 			}
-			_ => source_object_meta_inner.into_owned().headers,
+			_ => {
+				// The x-amz-website-redirect-location header is not copied, instead
+				// it is replaced by the value from the request (or removed if no
+				// value was specified)
+				let is_redirect =
+					|(key, _): &(String, String)| key == X_AMZ_WEBSITE_REDIRECT_LOCATION.as_str();
+				let mut headers: Vec<_> = source_object_meta_inner.headers.clone();
+				headers.retain(|h| !is_redirect(h));
+				let new_headers = extract_metadata_headers(req.headers())?;
+				headers.extend(new_headers.into_iter().filter(is_redirect));
+				headers
+			}
 		},
 		checksum: source_checksum,
+		checksum_type: source_checksum_type,
 	};
 
 	// Do actual object copying
@@ -116,8 +144,8 @@ pub async fn handle_copy(
 	// See: https://docs.aws.amazon.com/AmazonS3/latest/userguide/checking-object-integrity.html
 
 	let must_recopy = !EncryptionParams::is_same(&source_encryption, &dest_encryption)
-		|| source_checksum_algorithm != checksum_algorithm
-		|| (was_multipart && checksum_algorithm.is_some());
+		|| (checksum_algorithm.is_some()
+			&& (was_multipart || checksum_algorithm != source_checksum_algorithm));
 
 	let res = if !must_recopy {
 		// In most cases, we can just copy the metadata and link blocks of the
@@ -134,18 +162,21 @@ pub async fn handle_copy(
 		)
 		.await?
 	} else {
-		let expected_checksum = ExpectedChecksums {
-			md5: None,
-			sha256: None,
-			extra: source_checksum,
-		};
 		let checksum_mode = if was_multipart || source_checksum_algorithm != checksum_algorithm {
 			ChecksumMode::Calculate(checksum_algorithm)
 		} else {
-			ChecksumMode::Verify(&expected_checksum)
+			ChecksumMode::Verify(ExpectedChecksums {
+				md5: None,
+				sha256: None,
+				extra: source_checksum,
+			})
+		};
+		// For multipart uploads that had a composite checksum, set checksum type
+		// to full object as it will be recalculated.
+		let dest_object_meta = ObjectVersionMetaInner {
+			checksum_type: checksum_algorithm.map(|_| ChecksumType::FullObject),
+			..dest_object_meta
 		};
-		// If source and dest encryption use different keys,
-		// we must decrypt content and re-encrypt, so rewrite all data blocks.
 		handle_copy_reencrypt(
 			ctx,
 			dest_key,
@@ -237,6 +268,7 @@ async fn handle_copy_metaonly(
 				.get(&source_version.uuid, &EmptyKey)
 				.await?;
 			let source_version = source_version.ok_or(Error::NoSuchKey)?;
+			check_version_not_deleted(&source_version)?;
 
 			// Write an "uploading" marker in Object table
 			// This holds a reference to the object in the Version table
@@ -325,7 +357,7 @@ async fn handle_copy_reencrypt(
 	source_version: &ObjectVersion,
 	source_version_data: &ObjectVersionData,
 	source_encryption: EncryptionParams,
-	checksum_mode: ChecksumMode<'_>,
+	checksum_mode: ChecksumMode,
 ) -> Result<SaveStreamResult, Error> {
 	// basically we will read the source data (decrypt if necessary)
 	// and save that in a new object (encrypt if necessary),
@@ -437,6 +469,7 @@ pub async fn handle_upload_part_copy(
 		.get(&source_object_version.uuid, &EmptyKey)
 		.await?
 		.ok_or(Error::NoSuchKey)?;
+	check_version_not_deleted(&source_version)?;
 
 	// We want to reuse blocks from the source version as much as possible.
 	// However, we still need to get the data from these blocks
@@ -512,7 +545,7 @@ pub async fn handle_upload_part_copy(
 
 	// Now, actually copy the blocks
 	let mut checksummer = Checksummer::init(&Default::default(), !dest_encryption.is_encrypted())
-		.add(dest_object_checksum_algorithm);
+		.add(dest_object_checksum_algorithm.map(|(algo, _)| algo));
 
 	// First, create a stream that is able to read the source blocks
 	// and extract the subrange if necessary.
@@ -568,6 +601,7 @@ pub async fn handle_upload_part_copy(
 
 	let mut current_offset = 0;
 	let mut next_block = defragmenter.next().await?;
+	let mut blocks_to_dup = dest_version.clone();
 
 	// TODO this could be optimized similarly to read_and_put_blocks
 	// low priority because uploadpartcopy is rarely used
@@ -597,8 +631,7 @@ pub async fn handle_upload_part_copy(
 			.unwrap()?;
 		checksummer = checksummer_updated;
 
-		dest_version.blocks.clear();
-		dest_version.blocks.put(
+		let (version_block_key, version_block) = (
 			VersionBlockKey {
 				part_number,
 				offset: current_offset,
@@ -610,40 +643,59 @@ pub async fn handle_upload_part_copy(
 		);
 		current_offset += data_len;
 
-		let block_ref = BlockRef {
-			block: final_hash,
-			version: dest_version_id,
-			deleted: false.into(),
+		let next = if let Some(final_data) = data_to_upload {
+			dest_version.blocks.clear();
+			dest_version.blocks.put(version_block_key, version_block);
+			let block_ref = BlockRef {
+				block: final_hash,
+				version: dest_version_id,
+				deleted: false.into(),
+			};
+			let (_, _, _, next) = futures::try_join!(
+				// Thing 1: if the block is not exactly a block that existed before,
+				// we need to insert that data as a new block.
+				garage.block_manager.rpc_put_block(
+					final_hash,
+					final_data,
+					dest_encryption.is_encrypted(),
+					None
+				),
+				// Thing 2: we need to insert the block in the version
+				garage.version_table.insert(&dest_version),
+				// Thing 3: we need to add a block reference
+				garage.block_ref_table.insert(&block_ref),
+				// Thing 4: we need to read the next block
+				defragmenter.next(),
+			)?;
+			next
+		} else {
+			blocks_to_dup.blocks.put(version_block_key, version_block);
+			defragmenter.next().await?
 		};
-
-		let (_, _, _, next) = futures::try_join!(
-			// Thing 1: if the block is not exactly a block that existed before,
-			// we need to insert that data as a new block.
-			async {
-				if let Some(final_data) = data_to_upload {
-					garage
-						.block_manager
-						.rpc_put_block(final_hash, final_data, dest_encryption.is_encrypted(), None)
-						.await
-				} else {
-					Ok(())
-				}
-			},
-			// Thing 2: we need to insert the block in the version
-			garage.version_table.insert(&dest_version),
-			// Thing 3: we need to add a block reference
-			garage.block_ref_table.insert(&block_ref),
-			// Thing 4: we need to read the next block
-			defragmenter.next(),
-		)?;
 		next_block = next;
 	}
 
 	assert_eq!(current_offset, source_range.length);
 
+	// Put the duplicated blocks into the version & block_refs tables
+	let block_refs_to_put = blocks_to_dup
+		.blocks
+		.items()
+		.iter()
+		.map(|b| BlockRef {
+			block: b.1.hash,
+			version: dest_version_id,
+			deleted: false.into(),
+		})
+		.collect::<Vec<_>>();
+	futures::try_join!(
+		garage.version_table.insert(&blocks_to_dup),
+		garage.block_ref_table.insert_many(&block_refs_to_put[..]),
+	)?;
+
 	let checksums = checksummer.finalize();
 	let etag = dest_encryption.etag_from_md5(&checksums.md5);
-	let checksum = checksums.extract(dest_object_checksum_algorithm);
+	let checksum = checksums.extract(dest_object_checksum_algorithm.map(|(algo, _)| algo));
 
 	// Put the part's ETag in the Versiontable
 	dest_mpu.parts.put(

src/api/s3/get.rs

@@ -19,12 +19,13 @@ use garage_net::stream::ByteStream;
 use garage_rpc::rpc_helper::OrderTag;
 use garage_table::EmptyKey;
 use garage_util::data::*;
-use garage_util::error::OkOrMessage;
+use garage_util::error::{Error as UtilError, OkOrMessage};
 
 use garage_model::garage::Garage;
 use garage_model::s3::object_table::*;
 use garage_model::s3::version_table::*;
 
+use garage_api_common::common_error::CommonError;
 use garage_api_common::helpers::*;
 use garage_api_common::signature::checksum::{add_checksum_response_headers, X_AMZ_CHECKSUM_MODE};
 
@@ -219,6 +220,7 @@ pub async fn handle_head_without_ctx(
 					.get(&object_version.uuid, &EmptyKey)
 					.await?
 					.ok_or(Error::NoSuchKey)?;
+				check_version_not_deleted(&version)?;
 
 				let (part_offset, part_end) =
 					calculate_part_bounds(&version, pn).ok_or(Error::InvalidPart)?;
@@ -373,6 +375,21 @@ pub async fn handle_get_without_ctx(
 	}
 }
 
+pub(crate) fn check_version_not_deleted(version: &Version) -> Result<(), Error> {
+	if version.deleted.get() {
+		// the version was deleted between when the object_table was consulted
+		// and now, this could mean the object was deleted, or overriden.
+		// Rather than say the key doesn't exist, return a transient error
+		// to signal the client to try again.
+		return Err(CommonError::InternalError(UtilError::Message(
+			"conflict/inconsistency between object and version state, version is deleted"
+				.to_string(),
+		))
+		.into());
+	}
+	Ok(())
+}
+
 async fn handle_get_full(
 	garage: Arc<Garage>,
 	version: &ObjectVersion,
@@ -439,6 +456,7 @@ pub fn full_object_byte_stream(
 						.ok_or_message("channel closed")?;
 
 					let version = version_fut.await.unwrap()?.ok_or(Error::NoSuchKey)?;
+					check_version_not_deleted(&version)?;
 					for (i, (_, vb)) in version.blocks.items().iter().enumerate().skip(1) {
 						let stream_block_i = encryption
 							.get_block(&garage, &vb.hash, Some(order_stream.order(i as u64)))
@@ -454,6 +472,14 @@ pub fn full_object_byte_stream(
 				{
 					Ok(()) => (),
 					Err(e) => {
+						// TODO i think this is a bad idea, we should log
+						// an error and stop there. If the error happens to
+						// be exactly the size of what hasn't been streamed
+						// yet, the client will see the request as a
+						// success
+						// instead truncating the output notify the client
+						// something happened with their download, so that
+						// they can retry it
 						let _ = tx.send(error_stream_item(e)).await;
 					}
 				}
@@ -505,7 +531,7 @@ async fn handle_get_range(
 				.get(&version.uuid, &EmptyKey)
 				.await?
 				.ok_or(Error::NoSuchKey)?;
-
+			check_version_not_deleted(&version)?;
 			let body =
 				body_from_blocks_range(garage, encryption, version.blocks.items(), begin, end);
 			Ok(resp_builder.body(body)?)
@@ -556,6 +582,8 @@ async fn handle_get_part(
 				.await?
 				.ok_or(Error::NoSuchKey)?;
 
+			check_version_not_deleted(&version)?;
+
 			let (begin, end) =
 				calculate_part_bounds(&version, part_number).ok_or(Error::InvalidPart)?;
 

src/api/s3/lifecycle.rs

@@ -27,7 +27,7 @@ pub async fn handle_get_lifecycle(ctx: ReqCtx) -> Result<Response<ResBody>, Erro
 			.body(string_body(xml))?)
 	} else {
 		Ok(Response::builder()
-			.status(StatusCode::NO_CONTENT)
+			.status(StatusCode::NOT_FOUND)
 			.body(empty_body())?)
 	}
 }

src/api/s3/list.rs

@@ -1002,6 +1002,7 @@ mod tests {
 					inner: ObjectVersionMetaInner {
 						headers: vec![],
 						checksum: None,
+						checksum_type: None,
 					},
 				},
 				checksum_algorithm: None,

src/api/s3/multipart.rs

@@ -1,14 +1,11 @@
 use std::collections::HashMap;
-use std::convert::{TryFrom, TryInto};
-use std::hash::Hasher;
+use std::convert::TryInto;
 use std::sync::Arc;
 
 use base64::prelude::*;
-use crc32c::Crc32cHasher as Crc32c;
-use crc32fast::Hasher as Crc32;
-use crc64fast_nvme::Digest as Crc64Nvme;
+use crc_fast::{CrcAlgorithm, Digest as CrcDigest};
 use futures::prelude::*;
-use hyper::{Request, Response};
+use hyper::{header::HeaderValue, HeaderMap, Request, Response};
 use md5::{Digest, Md5};
 use sha1::Sha1;
 use sha2::Sha256;
@@ -54,6 +51,7 @@ pub async fn handle_create_multipart_upload(
 	let meta = ObjectVersionMetaInner {
 		headers,
 		checksum: None,
+		checksum_type: None,
 	};
 
 	// Determine whether object should be encrypted, and if so the key
@@ -68,7 +66,10 @@ pub async fn handle_create_multipart_upload(
 	)?;
 	let object_encryption = encryption.encrypt_meta(meta)?;
 
-	let checksum_algorithm = request_checksum_algorithm(req.headers())?;
+	let checksum_algorithm = request_checksum_algorithm_and_type(
+		req.headers(),
+		request_checksum_algorithm(req.headers())?,
+	)?;
 
 	// Create object in object table
 	let object_version = ObjectVersion {
@@ -227,7 +228,7 @@ pub async fn handle_put_part(
 		MpuPart {
 			version: version_uuid,
 			etag: Some(etag.clone()),
-			checksum: checksums.extract(checksum_algorithm),
+			checksum: checksums.extract(checksum_algorithm.map(|(algo, _)| algo)),
 			size: Some(total_size),
 		},
 	);
@@ -289,6 +290,14 @@ pub async fn handle_complete_multipart_upload(
 	let (req_head, req_body) = req.into_parts();
 
 	let expected_checksum = request_checksum_value(&req_head.headers)?;
+	let req_checksum_algorithm = request_checksum_algorithm_and_type(
+		&req_head.headers,
+		expected_checksum.map(|x| x.algorithm()),
+	)?;
+	debug!(
+		"CompleteMultipartUpload expected checksum: {:?}, request checksum type: {:?}",
+		expected_checksum, req_checksum_algorithm
+	);
 
 	let body = req_body.collect().await?;
 
@@ -319,6 +328,17 @@ pub async fn handle_complete_multipart_upload(
 		} => (encryption, checksum_algorithm),
 		_ => unreachable!(),
 	};
+	debug!(
+		"CompleteMultipartUpload object checksum_algorithm: {:?}",
+		checksum_algorithm
+	);
+	if req_checksum_algorithm.is_some() && req_checksum_algorithm != checksum_algorithm {
+		return Err(Error::InvalidDigest(format!(
+            "checksum algorithm {:?} does not correspond to algorithm specified in CreateMultipartUpload {:?}",
+            req_checksum_algorithm,
+            checksum_algorithm
+        )));
+	}
 
 	// Check that part numbers are an increasing sequence.
 	// (it doesn't need to start at 1 nor to be a continuous sequence,
@@ -344,8 +364,7 @@ pub async fn handle_complete_multipart_upload(
 	for req_part in body_list_of_parts.iter() {
 		match have_parts.get(&req_part.part_number) {
 			Some(part) if part.etag.as_ref() == Some(&req_part.etag) && part.size.is_some() => {
-				// alternative version: if req_part.checksum.is_some() && part.checksum != req_part.checksum {
-				if part.checksum != req_part.checksum {
+				if req_part.checksum.is_some() && part.checksum != req_part.checksum {
 					return Err(Error::InvalidDigest(format!(
 						"Invalid checksum for part {}: in request = {:?}, uploaded part = {:?}",
 						req_part.part_number, req_part.checksum, part.checksum
@@ -404,7 +423,11 @@ pub async fn handle_complete_multipart_upload(
 	// https://teppen.io/2018/06/23/aws_s3_etags/
 	let mut checksummer = MultipartChecksummer::init(checksum_algorithm);
 	for part in parts.iter() {
-		checksummer.update(part.etag.as_ref().unwrap(), part.checksum)?;
+		checksummer.update(
+			part.etag.as_ref().unwrap(),
+			part.checksum,
+			part.size.unwrap(),
+		)?;
 	}
 	let (checksum_md5, checksum_extra) = checksummer.finalize();
 
@@ -440,6 +463,7 @@ pub async fn handle_complete_multipart_upload(
 			let new_meta = ObjectVersionMetaInner {
 				headers: meta.into_owned().headers,
 				checksum: checksum_extra,
+				checksum_type: checksum_algorithm.map(|(_, ty)| ty),
 			};
 			encryption.encrypt_meta(new_meta)?
 		}
@@ -494,6 +518,11 @@ pub async fn handle_complete_multipart_upload(
 			Some(ChecksumValue::Sha256(x)) => Some(s3_xml::Value(BASE64_STANDARD.encode(&x))),
 			_ => None,
 		},
+		checksum_type: match checksum_algorithm {
+			Some((_, ChecksumType::Composite)) => Some(s3_xml::Value(COMPOSITE.into())),
+			Some((_, ChecksumType::FullObject)) => Some(s3_xml::Value(FULL_OBJECT.into())),
+			None => None,
+		},
 	};
 	let xml = s3_xml::to_xml_with_header(&result)?;
 
@@ -649,40 +678,52 @@ fn parse_complete_multipart_upload_body(
 
 // ====== checksummer ====
 
+pub fn request_checksum_algorithm_and_type(
+	headers: &HeaderMap<HeaderValue>,
+	algo: Option<ChecksumAlgorithm>,
+) -> Result<Option<(ChecksumAlgorithm, ChecksumType)>, Error> {
+	match (headers.get(X_AMZ_CHECKSUM_TYPE), algo) {
+		(None, None) => Ok(None),
+		(None, Some(algo)) => {
+			let ty = match algo {
+				ChecksumAlgorithm::Crc64Nvme => ChecksumType::FullObject,
+				_ => ChecksumType::Composite,
+			};
+			Ok(Some((algo, ty)))
+		}
+		(Some(_), None) => Err(Error::bad_request(
+			"Cannot specify x-amz-checksum-type when no checksum algorithm is in use.",
+		)),
+		(Some(x), Some(algo)) => {
+			let checksum_type = match x.as_bytes() {
+				x if x == COMPOSITE.as_bytes() => ChecksumType::Composite,
+				x if x == FULL_OBJECT.as_bytes() => ChecksumType::FullObject,
+				_ => return Err(Error::bad_request("Invalid x-amz-checksum-type value")),
+			};
+			match (checksum_type, algo) {
+				(ChecksumType::Composite, ChecksumAlgorithm::Crc64Nvme)
+				| (ChecksumType::FullObject, ChecksumAlgorithm::Sha1)
+				| (ChecksumType::FullObject, ChecksumAlgorithm::Sha256) => Err(Error::bad_request(format!(
+					"checksum type {:?} is not supported for algorithm {:?}",
+					checksum_type, algo
+				))),
+				(ty, algo) => Ok(Some((algo, ty))),
+			}
+		}
+	}
+}
+
 #[derive(Default)]
 pub(crate) struct MultipartChecksummer {
 	pub md5: Md5,
 	pub extra: Option<MultipartExtraChecksummer>,
 }
 
-pub(crate) enum MultipartExtraChecksummer {
-	Crc32(Crc32),
-	Crc32c(Crc32c),
-	Crc64Nvme(Crc64Nvme),
-	Sha1(Sha1),
-	Sha256(Sha256),
-}
-
 impl MultipartChecksummer {
-	pub(crate) fn init(algo: Option<ChecksumAlgorithm>) -> Self {
+	pub(crate) fn init(algo: Option<(ChecksumAlgorithm, ChecksumType)>) -> Self {
 		Self {
 			md5: Md5::new(),
-			extra: match algo {
-				None => None,
-				Some(ChecksumAlgorithm::Crc32) => {
-					Some(MultipartExtraChecksummer::Crc32(Crc32::new()))
-				}
-				Some(ChecksumAlgorithm::Crc32c) => {
-					Some(MultipartExtraChecksummer::Crc32c(Crc32c::default()))
-				}
-				Some(ChecksumAlgorithm::Crc64Nvme) => {
-					Some(MultipartExtraChecksummer::Crc64Nvme(Crc64Nvme::default()))
-				}
-				Some(ChecksumAlgorithm::Sha1) => Some(MultipartExtraChecksummer::Sha1(Sha1::new())),
-				Some(ChecksumAlgorithm::Sha256) => {
-					Some(MultipartExtraChecksummer::Sha256(Sha256::new()))
-				}
-			},
+			extra: algo.map(|(algo, cktype)| MultipartExtraChecksummer::init(algo, cktype)),
 		}
 	}
 
@@ -690,68 +731,130 @@ impl MultipartChecksummer {
 		&mut self,
 		etag: &str,
 		checksum: Option<ChecksumValue>,
+		part_len: u64,
 	) -> Result<(), Error> {
 		self.md5
 			.update(&hex::decode(&etag).ok_or_message("invalid etag hex")?);
-		match (&mut self.extra, checksum) {
-			(None, _) => (),
-			(
-				Some(MultipartExtraChecksummer::Crc32(ref mut crc32)),
-				Some(ChecksumValue::Crc32(x)),
-			) => {
-				crc32.update(&x);
-			}
-			(
-				Some(MultipartExtraChecksummer::Crc32c(ref mut crc32c)),
-				Some(ChecksumValue::Crc32c(x)),
-			) => {
-				crc32c.write(&x);
-			}
-			(
-				Some(MultipartExtraChecksummer::Crc64Nvme(ref mut crc64nvme)),
-				Some(ChecksumValue::Crc64Nvme(x)),
-			) => {
-				crc64nvme.write(&x);
-			}
-			(Some(MultipartExtraChecksummer::Sha1(ref mut sha1)), Some(ChecksumValue::Sha1(x))) => {
-				sha1.update(&x);
-			}
-			(
-				Some(MultipartExtraChecksummer::Sha256(ref mut sha256)),
-				Some(ChecksumValue::Sha256(x)),
-			) => {
-				sha256.update(&x);
-			}
-			(Some(_), b) => {
-				return Err(Error::internal_error(format!(
-					"part checksum was not computed correctly, got: {:?}",
-					b
-				)))
-			}
+		if let Some(extra) = &mut self.extra {
+			extra.update(checksum, part_len)?;
 		}
 		Ok(())
 	}
 
 	pub(crate) fn finalize(self) -> (Md5Checksum, Option<ChecksumValue>) {
 		let md5 = self.md5.finalize()[..].try_into().unwrap();
-		let extra = match self.extra {
-			None => None,
-			Some(MultipartExtraChecksummer::Crc32(crc32)) => {
-				Some(ChecksumValue::Crc32(u32::to_be_bytes(crc32.finalize())))
-			}
-			Some(MultipartExtraChecksummer::Crc32c(crc32c)) => Some(ChecksumValue::Crc32c(
-				u32::to_be_bytes(u32::try_from(crc32c.finish()).unwrap()),
-			)),
-			Some(MultipartExtraChecksummer::Crc64Nvme(crc64nvme)) => Some(
-				ChecksumValue::Crc64Nvme(u64::to_be_bytes(crc64nvme.sum64())),
-			),
-			Some(MultipartExtraChecksummer::Sha1(sha1)) => {
-				Some(ChecksumValue::Sha1(sha1.finalize()[..].try_into().unwrap()))
-			}
-			Some(MultipartExtraChecksummer::Sha256(sha256)) => Some(ChecksumValue::Sha256(
-				sha256.finalize()[..].try_into().unwrap(),
-			)),
-		};
+		let extra = self.extra.map(|c| c.finalize());
 		(md5, extra)
 	}
 }
+
+pub(crate) enum MultipartExtraChecksummer {
+	FullObjectCrc(CrcAlgorithm, Option<u64>),
+	CompositeCrc(ChecksumAlgorithm, CrcDigest),
+	CompositeSha1(Sha1),
+	CompositeSha256(Sha256),
+}
+
+impl MultipartExtraChecksummer {
+	fn init(algo: ChecksumAlgorithm, cktype: ChecksumType) -> Self {
+		match (algo, cktype) {
+			(algo, ChecksumType::FullObject) => {
+				let crc_type = match algo {
+					ChecksumAlgorithm::Crc32 => CrcAlgorithm::Crc32IsoHdlc,
+					ChecksumAlgorithm::Crc32c => CrcAlgorithm::Crc32Iscsi,
+					ChecksumAlgorithm::Crc64Nvme => CrcAlgorithm::Crc64Nvme,
+					_ => unreachable!(),
+				};
+				Self::FullObjectCrc(crc_type, None)
+			}
+			(ChecksumAlgorithm::Crc32, ChecksumType::Composite) => {
+				Self::CompositeCrc(ChecksumAlgorithm::Crc32, new_crc32())
+			}
+			(ChecksumAlgorithm::Crc32c, ChecksumType::Composite) => {
+				Self::CompositeCrc(ChecksumAlgorithm::Crc32c, new_crc32c())
+			}
+			(ChecksumAlgorithm::Sha1, ChecksumType::Composite) => Self::CompositeSha1(Sha1::new()),
+			(ChecksumAlgorithm::Sha256, ChecksumType::Composite) => {
+				Self::CompositeSha256(Sha256::new())
+			}
+			_ => unreachable!(),
+		}
+	}
+
+	fn update(&mut self, checksum: Option<ChecksumValue>, part_len: u64) -> Result<(), Error> {
+		match (self, checksum) {
+			(Self::FullObjectCrc(crc_algo, crc_value), Some(ck)) => {
+				let ck_u64 = match ck {
+					ChecksumValue::Crc32(x) => u32::from_be_bytes(x) as u64,
+					ChecksumValue::Crc32c(x) => u32::from_be_bytes(x) as u64,
+					ChecksumValue::Crc64Nvme(x) => u64::from_be_bytes(x),
+					_ => {
+						return Err(Error::internal_error(format!(
+							"part checksum was not computed correctly, got: {:?}",
+							ck
+						)))
+					}
+				};
+				*crc_value = match *crc_value {
+					None => Some(ck_u64),
+					Some(prev) => Some(crc_fast::checksum_combine(
+						*crc_algo, prev, ck_u64, part_len,
+					)),
+				};
+			}
+			(Self::CompositeCrc(_, digest), Some(ck)) => match ck {
+				ChecksumValue::Crc32(x) => digest.update(&x),
+				ChecksumValue::Crc32c(x) => digest.update(&x),
+				ChecksumValue::Crc64Nvme(x) => digest.update(&x),
+				_ => {
+					return Err(Error::internal_error(format!(
+						"part checksum was not computed correctly, got: {:?}",
+						ck
+					)))
+				}
+			},
+			(Self::CompositeSha1(sha1), Some(ChecksumValue::Sha1(x))) => {
+				sha1.update(&x);
+			}
+			(Self::CompositeSha256(sha256), Some(ChecksumValue::Sha256(x))) => {
+				sha256.update(&x);
+			}
+			_ => {
+				return Err(Error::internal_error(format!(
+					"part checksum was not computed correctly, got: {:?}",
+					checksum
+				)))
+			}
+		}
+		Ok(())
+	}
+	fn finalize(self) -> ChecksumValue {
+		match self {
+			Self::FullObjectCrc(algo, value) => match (algo, value) {
+				(CrcAlgorithm::Crc32IsoHdlc, Some(v)) => {
+					ChecksumValue::Crc32(u32::to_be_bytes(v as u32))
+				}
+				(CrcAlgorithm::Crc32Iscsi, Some(v)) => {
+					ChecksumValue::Crc32c(u32::to_be_bytes(v as u32))
+				}
+				(CrcAlgorithm::Crc64Nvme, Some(v)) => ChecksumValue::Crc64Nvme(u64::to_be_bytes(v)),
+				_ => unreachable!(),
+			},
+			Self::CompositeCrc(algo, crc) => match algo {
+				ChecksumAlgorithm::Crc32 => {
+					ChecksumValue::Crc32(u32::to_be_bytes(crc.finalize() as u32))
+				}
+				ChecksumAlgorithm::Crc32c => {
+					ChecksumValue::Crc32c(u32::to_be_bytes(crc.finalize() as u32))
+				}
+				_ => unreachable!(),
+			},
+			Self::CompositeSha1(sha1) => {
+				ChecksumValue::Sha1(sha1.finalize()[..].try_into().unwrap())
+			}
+			Self::CompositeSha256(sha256) => {
+				ChecksumValue::Sha256(sha256.finalize()[..].try_into().unwrap())
+			}
+		}
+	}
+}

src/api/s3/post_object.rs

@@ -233,6 +233,7 @@ pub async fn handle_post_object(
 	let meta = ObjectVersionMetaInner {
 		headers,
 		checksum: expected_checksums.extra,
+		checksum_type: expected_checksums.extra.map(|_| ChecksumType::FullObject),
 	};
 
 	let encryption = EncryptionParams::new_from_headers(
@@ -261,7 +262,7 @@ pub async fn handle_post_object(
 		encryption,
 		StreamLimiter::new(stream, conditions.content_length),
 		&key,
-		ChecksumMode::Verify(&expected_checksums),
+		ChecksumMode::Verify(expected_checksums),
 	)
 	.await?;
 

src/api/s3/put.rs

@@ -48,8 +48,8 @@ pub(crate) struct SaveStreamResult {
 	pub(crate) etag: String,
 }
 
-pub(crate) enum ChecksumMode<'a> {
-	Verify(&'a ExpectedChecksums),
+pub(crate) enum ChecksumMode {
+	Verify(ExpectedChecksums),
 	VerifyFrom {
 		checksummer: StreamingChecksumReceiver,
 		trailer_algo: Option<ChecksumAlgorithm>,
@@ -83,6 +83,7 @@ pub async fn handle_put(
 	let meta = ObjectVersionMetaInner {
 		headers,
 		checksum: expected_checksums.extra,
+		checksum_type: expected_checksums.extra.map(|_| ChecksumType::FullObject),
 	};
 
 	// Determine whether object should be encrypted, and if so the key
@@ -139,7 +140,7 @@ pub(crate) async fn save_stream<S: Stream<Item = Result<Bytes, Error>> + Unpin>(
 	encryption: EncryptionParams,
 	body: S,
 	key: &String,
-	checksum_mode: ChecksumMode<'_>,
+	checksum_mode: ChecksumMode,
 ) -> Result<SaveStreamResult, Error> {
 	let ReqCtx {
 		garage, bucket_id, ..

src/api/s3/xml.rs

@@ -13,6 +13,10 @@ pub fn xmlns_tag<S: Serializer>(_v: &(), s: S) -> Result<S::Ok, S::Error> {
 	s.serialize_str("http://s3.amazonaws.com/doc/2006-03-01/")
 }
 
+pub fn xmlns_xsi_tag<S: Serializer>(_v: &(), s: S) -> Result<S::Ok, S::Error> {
+	s.serialize_str("http://www.w3.org/2001/XMLSchema-instance")
+}
+
 #[derive(Debug, Serialize, Deserialize, PartialEq, Eq, PartialOrd, Ord)]
 pub struct Value(#[serde(rename = "$value")] pub String);
 
@@ -141,6 +145,8 @@ pub struct CompleteMultipartUploadResult {
 	pub checksum_sha1: Option<Value>,
 	#[serde(rename = "ChecksumSHA256")]
 	pub checksum_sha256: Option<Value>,
+	#[serde(rename = "ChecksumType")]
+	pub checksum_type: Option<Value>,
 }
 
 #[derive(Debug, Serialize, PartialEq, Eq)]
@@ -323,6 +329,42 @@ pub struct PostObject {
 	pub etag: Value,
 }
 
+#[derive(Debug, Serialize, PartialEq, Eq)]
+pub struct Grantee {
+	#[serde(rename = "xmlns:xsi", serialize_with = "xmlns_xsi_tag")]
+	pub xmlns_xsi: (),
+	#[serde(rename = "xsi:type")]
+	pub typ: String,
+	#[serde(rename = "DisplayName")]
+	pub display_name: Option<Value>,
+	#[serde(rename = "ID")]
+	pub id: Option<Value>,
+}
+
+#[derive(Debug, Serialize, PartialEq, Eq)]
+pub struct Grant {
+	#[serde(rename = "Grantee")]
+	pub grantee: Grantee,
+	#[serde(rename = "Permission")]
+	pub permission: Value,
+}
+
+#[derive(Debug, Serialize, PartialEq, Eq)]
+pub struct AccessControlList {
+	#[serde(rename = "Grant")]
+	pub entries: Vec<Grant>,
+}
+
+#[derive(Debug, Serialize, PartialEq, Eq)]
+pub struct AccessControlPolicy {
+	#[serde(serialize_with = "xmlns_tag")]
+	pub xmlns: (),
+	#[serde(rename = "Owner")]
+	pub owner: Option<Owner>,
+	#[serde(rename = "AccessControlList")]
+	pub acl: AccessControlList,
+}
+
 #[cfg(test)]
 mod tests {
 	use super::*;
@@ -431,6 +473,43 @@ mod tests {
 		Ok(())
 	}
 
+	#[test]
+	fn get_bucket_acl_result() -> Result<(), ApiError> {
+		let grant = Grant {
+			grantee: Grantee {
+				xmlns_xsi: (),
+				typ: "CanonicalUser".to_string(),
+				display_name: Some(Value("owner_name".to_string())),
+				id: Some(Value("qsdfjklm".to_string())),
+			},
+			permission: Value("FULL_CONTROL".to_string()),
+		};
+
+		let get_bucket_acl = AccessControlPolicy {
+			xmlns: (),
+			owner: None,
+			acl: AccessControlList {
+				entries: vec![grant],
+			},
+		};
+		assert_eq!(
+			to_xml_with_header(&get_bucket_acl)?,
+			"<?xml version=\"1.0\" encoding=\"UTF-8\"?>\
+<AccessControlPolicy xmlns=\"http://s3.amazonaws.com/doc/2006-03-01/\">\
+	<AccessControlList>\
+		<Grant>\
+			<Grantee xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\" xsi:type=\"CanonicalUser\">\
+				<DisplayName>owner_name</DisplayName>\
+				<ID>qsdfjklm</ID>\
+			</Grantee>\
+			<Permission>FULL_CONTROL</Permission>\
+		</Grant>\
+	</AccessControlList>\
+</AccessControlPolicy>"
+		);
+		Ok(())
+	}
+
 	#[test]
 	fn delete_result() -> Result<(), ApiError> {
 		let delete_result = DeleteResult {
@@ -514,6 +593,7 @@ mod tests {
 
 	#[test]
 	fn complete_multipart_upload_result() -> Result<(), ApiError> {
+		use garage_api_common::signature::checksum::COMPOSITE;
 		let result = CompleteMultipartUploadResult {
 			xmlns: (),
 			location: Some(Value("https://garage.tld/mybucket/a/plop".to_string())),
@@ -525,6 +605,7 @@ mod tests {
 			checksum_crc64nvme: None,
 			checksum_sha1: Some(Value("ZJAnHyG8PeKz9tI8UTcHrJos39A=".into())),
 			checksum_sha256: None,
+			checksum_type: Some(Value(COMPOSITE.into())),
 		};
 		assert_eq!(
 			to_xml_with_header(&result)?,
@@ -535,6 +616,7 @@ mod tests {
 	<Key>a/plop</Key>\
 	<ETag>&quot;3858f62230ac3c915f300c664312c11f-9&quot;</ETag>\
     <ChecksumSHA1>ZJAnHyG8PeKz9tI8UTcHrJos39A=</ChecksumSHA1>\
+    <ChecksumType>COMPOSITE</ChecksumType>\
 </CompleteMultipartUploadResult>"
 		);
 		Ok(())

src/db/lmdb_adapter.rs

@@ -109,7 +109,7 @@ impl IDb for LmdbDb {
 		let mut path = to.clone();
 		path.push("data.mdb");
 		self.db
-			.copy_to_path(path, heed::CompactionOption::Disabled)?;
+			.copy_to_path(path, heed::CompactionOption::Enabled)?;
 		Ok(())
 	}
 

src/garage/Cargo.toml

@@ -59,6 +59,7 @@ opentelemetry.workspace = true
 opentelemetry-prometheus = { workspace = true, optional = true }
 opentelemetry-otlp = { workspace = true, optional = true }
 syslog-tracing = { workspace = true, optional = true }
+tracing-journald = { workspace = true, optional = true }
 
 [dev-dependencies]
 garage_api_common.workspace = true
@@ -79,7 +80,7 @@ static_init.workspace = true
 assert-json-diff.workspace = true
 serde_json.workspace = true
 base64.workspace = true
-crc32fast.workspace = true
+crc-fast.workspace = true
 
 k2v-client.workspace = true
 
@@ -103,6 +104,8 @@ metrics = [ "garage_api_admin/metrics", "opentelemetry-prometheus" ]
 telemetry-otlp = [ "opentelemetry-otlp" ]
 # Logging to syslog
 syslog = [ "syslog-tracing" ]
+# Logging to journald
+journald = [ "tracing-journald" ]
 
 # NOTE: bundled-libs and system-libs should be treat as mutually exclusive;
 # exactly one of them should be enabled.

src/garage/cli/remote/block.rs

@@ -160,8 +160,8 @@ impl Cli {
 			.await?;
 
 		println!(
-			"Purged {} blocks: deleted {} versions, {} objects, {} multipart uploads",
-			res.blocks_purged, res.versions_deleted, res.objects_deleted, res.uploads_deleted,
+			"Purged {} blocks: deleted {} block refs, {} versions, {} objects, {} multipart uploads",
+			res.blocks_purged, res.block_refs_purged, res.versions_deleted, res.objects_deleted, res.uploads_deleted,
 		);
 
 		Ok(())

src/garage/cli/remote/node.rs

@@ -91,6 +91,7 @@ impl Cli {
 				ScrubCmd::Pause => ScrubCommand::Pause,
 				ScrubCmd::Resume => ScrubCommand::Resume,
 			}),
+			RepairWhat::Aliases => RepairType::Aliases,
 		};
 
 		let res = self

src/garage/cli/structs.rs

@@ -657,6 +657,9 @@ pub enum RepairWhat {
 	/// Recalculate block reference counters
 	#[structopt(name = "block-rc", version = garage_version())]
 	BlockRc,
+	/// Fix inconsistency in bucket aliases (WARNING: EXPERIMENTAL)
+	#[structopt(name = "aliases", version = garage_version())]
+	Aliases,
 	/// Verify integrity of all blocks on disc
 	#[structopt(name = "scrub", version = garage_version())]
 	Scrub {

src/garage/main.rs

@@ -215,6 +215,43 @@ fn init_logging(opt: &Opt) {
 		}
 	}
 
+	if std::env::var("GARAGE_LOG_TO_JOURNALD")
+		.map(|x| x == "1" || x == "true")
+		.unwrap_or(false)
+	{
+		#[cfg(feature = "journald")]
+		{
+			use tracing_journald::{Priority, PriorityMappings};
+			use tracing_subscriber::layer::SubscriberExt;
+			use tracing_subscriber::util::SubscriberInitExt;
+
+			let registry = tracing_subscriber::registry()
+				.with(tracing_subscriber::fmt::layer().with_writer(std::io::sink))
+				.with(env_filter);
+			match tracing_journald::layer() {
+				Ok(layer) => {
+					registry
+						.with(layer.with_priority_mappings(PriorityMappings {
+							info: Priority::Informational,
+							debug: Priority::Debug,
+							..PriorityMappings::new()
+						}))
+						.init();
+				}
+				Err(e) => {
+					eprintln!("Couldn't connect to journald: {}.", e);
+					std::process::exit(1);
+				}
+			}
+			return;
+		}
+		#[cfg(not(feature = "journald"))]
+		{
+			eprintln!("Journald support is not enabled in this build.");
+			std::process::exit(1);
+		}
+	}
+
 	tracing_subscriber::fmt()
 		.with_writer(std::io::stderr)
 		.with_env_filter(env_filter)

src/garage/tests/common/garage.rs

@@ -65,6 +65,8 @@ rpc_bind_addr = "127.0.0.1:{rpc_port}"
 rpc_public_addr = "127.0.0.1:{rpc_port}"
 rpc_secret = "{secret}"
 
+allow_punycode = true
+
 [s3_api]
 s3_region = "{region}"
 api_bind_addr = "127.0.0.1:{s3_port}"

src/garage/tests/s3/streaming_signature.rs

@@ -1,7 +1,7 @@
 use std::collections::HashMap;
 
 use base64::prelude::*;
-use crc32fast::Hasher as Crc32;
+use crc_fast::{checksum as crc_checksum, CrcAlgorithm};
 
 use crate::common;
 use crate::common::ext::CommandExt;
@@ -69,9 +69,8 @@ async fn test_putobject_streaming() {
 	{
 		let etag = "\"46cf18a9b447991b450cad3facf5937e\"";
 
-		let mut crc32 = Crc32::new();
-		crc32.update(&BODY[..]);
-		let crc32 = BASE64_STANDARD.encode(&u32::to_be_bytes(crc32.finalize())[..]);
+		let crc32 = crc_checksum(CrcAlgorithm::Crc32IsoHdlc, &BODY[..]) as u32;
+		let crc32 = BASE64_STANDARD.encode(&u32::to_be_bytes(crc32)[..]);
 
 		let mut headers = HashMap::new();
 		headers.insert("x-amz-checksum-crc32".to_owned(), crc32.clone());
@@ -129,7 +128,8 @@ async fn test_putobject_streaming_unsigned_trailer() {
 		let mut headers = HashMap::new();
 		headers.insert("content-type".to_owned(), content_type.to_owned());
 
-		let empty_crc32 = BASE64_STANDARD.encode(&u32::to_be_bytes(Crc32::new().finalize())[..]);
+		let empty_crc32 = crc_checksum(CrcAlgorithm::Crc32IsoHdlc, &[]) as u32;
+		let empty_crc32 = BASE64_STANDARD.encode(&u32::to_be_bytes(empty_crc32)[..]);
 
 		let res = ctx
 			.custom_request
@@ -180,9 +180,8 @@ async fn test_putobject_streaming_unsigned_trailer() {
 	{
 		let etag = "\"46cf18a9b447991b450cad3facf5937e\"";
 
-		let mut crc32 = Crc32::new();
-		crc32.update(&BODY[..]);
-		let crc32 = BASE64_STANDARD.encode(&u32::to_be_bytes(crc32.finalize())[..]);
+		let crc32 = crc_checksum(CrcAlgorithm::Crc32IsoHdlc, &BODY[..]) as u32;
+		let crc32 = BASE64_STANDARD.encode(&u32::to_be_bytes(crc32)[..]);
 
 		// try sending with wrong crc32, check that it fails
 		let err_res = ctx

src/garage/tests/s3/website.rs

@@ -984,3 +984,118 @@ async fn test_website_invalid_redirect() {
 		.await
 		.unwrap_err();
 }
+
+#[tokio::test]
+async fn test_website_puny() {
+	const BCKT_NAME: &str = "xn--pda.eu";
+	let ctx = common::context();
+	let bucket = ctx.create_bucket(BCKT_NAME);
+
+	let data = ByteStream::from_static(BODY);
+
+	ctx.client
+		.put_object()
+		.bucket(&bucket)
+		.key("index.html")
+		.body(data)
+		.send()
+		.await
+		.unwrap();
+
+	let client = Client::builder(TokioExecutor::new()).build_http();
+
+	let req = |suffix| {
+		Request::builder()
+			.method("GET")
+			.uri(format!("http://127.0.0.1:{}/", ctx.garage.web_port))
+			.header("Host", format!("{}{}", BCKT_NAME, suffix))
+			.body(Body::new(Bytes::new()))
+			.unwrap()
+	};
+
+	ctx.garage
+		.command()
+		.args(["bucket", "website", "--allow", BCKT_NAME])
+		.quiet()
+		.expect_success_status("Could not allow website on bucket");
+
+	let mut resp = client.request(req("")).await.unwrap();
+	assert_eq!(resp.status(), StatusCode::OK);
+	assert_eq!(
+		resp.into_body().collect().await.unwrap().to_bytes(),
+		BODY.as_ref()
+	);
+
+	resp = client.request(req(".web.garage")).await.unwrap();
+	assert_eq!(resp.status(), StatusCode::OK);
+	assert_eq!(
+		resp.into_body().collect().await.unwrap().to_bytes(),
+		BODY.as_ref()
+	);
+
+	for bname in [
+		BCKT_NAME.to_string(),
+		format!("{BCKT_NAME}.web.garage"),
+		format!("{BCKT_NAME}.s3.garage"),
+	] {
+		let admin_req = || {
+			Request::builder()
+				.method("GET")
+				.uri(format!(
+					"http://127.0.0.1:{0}/check?domain={1}",
+					ctx.garage.admin_port, bname
+				))
+				.body(Body::new(Bytes::new()))
+				.unwrap()
+		};
+
+		let admin_resp = client.request(admin_req()).await.unwrap();
+		assert_eq!(admin_resp.status(), StatusCode::OK);
+		assert_eq!(
+			admin_resp.into_body().collect().await.unwrap().to_bytes(),
+			format!("Domain '{bname}' is managed by Garage").as_bytes()
+		);
+	}
+}
+
+#[tokio::test]
+async fn test_website_object_not_found() {
+	const BCKT_NAME: &str = "not-found";
+	let ctx = common::context();
+	let _bucket = ctx.create_bucket(BCKT_NAME);
+
+	let client = Client::builder(TokioExecutor::new()).build_http();
+
+	let req = |suffix| {
+		Request::builder()
+			.method("GET")
+			.uri(format!("http://127.0.0.1:{}/", ctx.garage.web_port))
+			.header("Host", format!("{}{}", BCKT_NAME, suffix))
+			.body(Body::new(Bytes::new()))
+			.unwrap()
+	};
+
+	ctx.garage
+		.command()
+		.args(["bucket", "website", "--allow", BCKT_NAME])
+		.quiet()
+		.expect_success_status("Could not allow website on bucket");
+
+	let resp = client.request(req("")).await.unwrap();
+	assert_eq!(resp.status(), StatusCode::NOT_FOUND);
+	// the error we return by default are *not* xml
+	assert_eq!(
+		resp.headers().get(http::header::CONTENT_TYPE).unwrap(),
+		"text/html; charset=utf-8"
+	);
+	let result = String::from_utf8(
+		resp.into_body()
+			.collect()
+			.await
+			.unwrap()
+			.to_bytes()
+			.to_vec(),
+	)
+	.unwrap();
+	assert!(result.contains("not found"));
+}

src/model/bucket_alias_table.rs

@@ -22,14 +22,10 @@ mod v08 {
 pub use v08::*;
 
 impl BucketAlias {
-	pub fn new(name: String, ts: u64, bucket_id: Option<Uuid>) -> Option<Self> {
-		if !is_valid_bucket_name(&name) {
-			None
-		} else {
-			Some(BucketAlias {
-				name,
-				state: crdt::Lww::raw(ts, bucket_id),
-			})
+	pub fn new(name: String, ts: u64, bucket_id: Option<Uuid>) -> Self {
+		BucketAlias {
+			name,
+			state: crdt::Lww::raw(ts, bucket_id),
 		}
 	}
 
@@ -80,7 +76,7 @@ impl TableSchema for BucketAliasTable {
 /// In the case of Garage, bucket names must not be hex-encoded
 /// 32 byte string, which is excluded thanks to the
 /// maximum length of 63 bytes given in the spec.
-pub fn is_valid_bucket_name(n: &str) -> bool {
+pub fn is_valid_bucket_name(n: &str, puny: bool) -> bool {
 	// Bucket names must be between 3 and 63 characters
 	n.len() >= 3 && n.len() <= 63
 	// Bucket names must be composed of lowercase letters, numbers,
@@ -92,7 +88,9 @@ pub fn is_valid_bucket_name(n: &str) -> bool {
 	// Bucket names must not be formatted as an IP address
 	&& n.parse::<std::net::IpAddr>().is_err()
 	// Bucket names must not start with "xn--"
-	&& !n.starts_with("xn--")
+	&& (!n.starts_with("xn--") || puny)
+        // We are a bit stricter, to properly restrict punycode in all labels
+	&& (!n.contains(".xn--") || puny)
 	// Bucket names must not end with "-s3alias"
 	&& !n.ends_with("-s3alias")
 }

src/model/helper/locked.rs

@@ -1,3 +1,7 @@
+use std::collections::{HashMap, HashSet};
+
+use garage_db as db;
+
 use garage_util::crdt::*;
 use garage_util::data::*;
 use garage_util::error::{Error as GarageError, OkOrMessage};
@@ -47,6 +51,10 @@ impl<'a> LockedHelper<'a> {
 		KeyHelper(self.0)
 	}
 
+	// ================================================
+	//      global bucket aliases
+	// ================================================
+
 	/// Sets a new alias for a bucket in global namespace.
 	/// This function fails if:
 	/// - alias name is not valid according to S3 spec
@@ -57,7 +65,7 @@ impl<'a> LockedHelper<'a> {
 		bucket_id: Uuid,
 		alias_name: &String,
 	) -> Result<(), Error> {
-		if !is_valid_bucket_name(alias_name) {
+		if !is_valid_bucket_name(alias_name, self.0.config.allow_punycode) {
 			return Err(Error::InvalidBucketName(alias_name.to_string()));
 		}
 
@@ -88,8 +96,7 @@ impl<'a> LockedHelper<'a> {
 		// writes are now done and all writes use timestamp alias_ts
 
 		let alias = match alias {
-			None => BucketAlias::new(alias_name.clone(), alias_ts, Some(bucket_id))
-				.ok_or_else(|| Error::InvalidBucketName(alias_name.clone()))?,
+			None => BucketAlias::new(alias_name.clone(), alias_ts, Some(bucket_id)),
 			Some(mut a) => {
 				a.state = Lww::raw(alias_ts, Some(bucket_id));
 				a
@@ -180,13 +187,14 @@ impl<'a> LockedHelper<'a> {
 			.ok_or_else(|| Error::NoSuchBucket(alias_name.to_string()))?;
 
 		// Checks ok, remove alias
-		let alias_ts = match bucket.state.as_option() {
-			Some(bucket_state) => increment_logical_clock_2(
-				alias.state.timestamp(),
-				bucket_state.aliases.get_timestamp(alias_name),
-			),
-			None => increment_logical_clock(alias.state.timestamp()),
-		};
+		let alias_ts = increment_logical_clock_2(
+			alias.state.timestamp(),
+			bucket
+				.state
+				.as_option()
+				.map(|p| p.aliases.get_timestamp(alias_name))
+				.unwrap_or(0),
+		);
 
 		// ---- timestamp-ensured causality barrier ----
 		// writes are now done and all writes use timestamp alias_ts
@@ -204,6 +212,10 @@ impl<'a> LockedHelper<'a> {
 		Ok(())
 	}
 
+	// ================================================
+	//      local bucket aliases
+	// ================================================
+
 	/// Sets a new alias for a bucket in the local namespace of a key.
 	/// This function fails if:
 	/// - alias name is not valid according to S3 spec
@@ -216,14 +228,12 @@ impl<'a> LockedHelper<'a> {
 		key_id: &String,
 		alias_name: &String,
 	) -> Result<(), Error> {
-		let key_helper = KeyHelper(self.0);
-
-		if !is_valid_bucket_name(alias_name) {
+		if !is_valid_bucket_name(alias_name, self.0.config.allow_punycode) {
 			return Err(Error::InvalidBucketName(alias_name.to_string()));
 		}
 
 		let mut bucket = self.bucket().get_existing_bucket(bucket_id).await?;
-		let mut key = key_helper.get_existing_key(key_id).await?;
+		let mut key = self.key().get_existing_key(key_id).await?;
 
 		let key_param = key.state.as_option_mut().unwrap();
 
@@ -272,23 +282,13 @@ impl<'a> LockedHelper<'a> {
 		key_id: &String,
 		alias_name: &String,
 	) -> Result<(), Error> {
-		let key_helper = KeyHelper(self.0);
-
 		let mut bucket = self.bucket().get_existing_bucket(bucket_id).await?;
-		let mut key = key_helper.get_existing_key(key_id).await?;
+		let mut key = self.key().get_existing_key(key_id).await?;
 
+		let key_p = key.state.as_option().unwrap();
 		let bucket_p = bucket.state.as_option_mut().unwrap();
 
-		if key
-			.state
-			.as_option()
-			.unwrap()
-			.local_aliases
-			.get(alias_name)
-			.cloned()
-			.flatten()
-			!= Some(bucket_id)
-		{
+		if key_p.local_aliases.get(alias_name).cloned().flatten() != Some(bucket_id) {
 			return Err(GarageError::Message(format!(
 				"Bucket {:?} does not have alias {} in namespace of key {}",
 				bucket_id, alias_name, key_id
@@ -305,17 +305,17 @@ impl<'a> LockedHelper<'a> {
 			.local_aliases
 			.items()
 			.iter()
-			.any(|((k, n), _, active)| *k == key.key_id && n == alias_name && *active);
+			.any(|((k, n), _, active)| (*k != key.key_id || n != alias_name) && *active);
+
 		if !has_other_global_aliases && !has_other_local_aliases {
 			return Err(Error::BadRequest(format!("Bucket {} doesn't have other aliases, please delete it instead of just unaliasing.", alias_name)));
 		}
 
 		// Checks ok, remove alias
-		let key_param = key.state.as_option_mut().unwrap();
 		let bucket_p_local_alias_key = (key.key_id.clone(), alias_name.clone());
 
 		let alias_ts = increment_logical_clock_2(
-			key_param.local_aliases.get_timestamp(alias_name),
+			key_p.local_aliases.get_timestamp(alias_name),
 			bucket_p
 				.local_aliases
 				.get_timestamp(&bucket_p_local_alias_key),
@@ -324,7 +324,8 @@ impl<'a> LockedHelper<'a> {
 		// ---- timestamp-ensured causality barrier ----
 		// writes are now done and all writes use timestamp alias_ts
 
-		key_param.local_aliases = LwwMap::raw_item(alias_name.clone(), alias_ts, None);
+		key.state.as_option_mut().unwrap().local_aliases =
+			LwwMap::raw_item(alias_name.clone(), alias_ts, None);
 		self.0.key_table.insert(&key).await?;
 
 		bucket_p.local_aliases = LwwMap::raw_item(bucket_p_local_alias_key, alias_ts, false);
@@ -333,21 +334,68 @@ impl<'a> LockedHelper<'a> {
 		Ok(())
 	}
 
+	/// Ensures a bucket does not have a certain local alias.
+	/// Contrarily to unset_local_bucket_alias, this does not
+	/// fail on any condition other than:
+	/// - bucket cannot be found (its fine if it is in deleted state)
+	/// - key cannot be found (its fine if alias in key points to nothing
+	///   or to another bucket)
+	pub async fn purge_local_bucket_alias(
+		&self,
+		bucket_id: Uuid,
+		key_id: &String,
+		alias_name: &String,
+	) -> Result<(), Error> {
+		let mut bucket = self.bucket().get_internal_bucket(bucket_id).await?;
+		let mut key = self.key().get_internal_key(key_id).await?;
+
+		let bucket_p_local_alias_key = (key.key_id.clone(), alias_name.clone());
+
+		let alias_ts = increment_logical_clock_2(
+			key.state
+				.as_option()
+				.map(|p| p.local_aliases.get_timestamp(alias_name))
+				.unwrap_or(0),
+			bucket
+				.state
+				.as_option()
+				.map(|p| p.local_aliases.get_timestamp(&bucket_p_local_alias_key))
+				.unwrap_or(0),
+		);
+
+		// ---- timestamp-ensured causality barrier ----
+		// writes are now done and all writes use timestamp alias_ts
+
+		if let Some(kp) = key.state.as_option_mut() {
+			kp.local_aliases = LwwMap::raw_item(alias_name.clone(), alias_ts, None);
+			self.0.key_table.insert(&key).await?;
+		}
+
+		if let Some(bp) = bucket.state.as_option_mut() {
+			bp.local_aliases = LwwMap::raw_item(bucket_p_local_alias_key, alias_ts, false);
+			self.0.bucket_table.insert(&bucket).await?;
+		}
+
+		Ok(())
+	}
+
+	// ================================================
+	//      permissions
+	// ================================================
+
 	/// Sets permissions for a key on a bucket.
 	/// This function fails if:
 	/// - bucket or key cannot be found at all (its ok if they are in deleted state)
-	/// - bucket or key is in deleted state and we are trying to set permissions other than "deny
-	/// all"
+	/// - bucket or key is in deleted state and we are trying to set
+	///   permissions other than "deny all"
 	pub async fn set_bucket_key_permissions(
 		&self,
 		bucket_id: Uuid,
 		key_id: &String,
 		mut perm: BucketKeyPerm,
 	) -> Result<(), Error> {
-		let key_helper = KeyHelper(self.0);
-
 		let mut bucket = self.bucket().get_internal_bucket(bucket_id).await?;
-		let mut key = key_helper.get_internal_key(key_id).await?;
+		let mut key = self.key().get_internal_key(key_id).await?;
 
 		if let Some(bstate) = bucket.state.as_option() {
 			if let Some(kp) = bstate.authorized_keys.get(key_id) {
@@ -384,21 +432,20 @@ impl<'a> LockedHelper<'a> {
 		Ok(())
 	}
 
-	// ----
+	// ================================================
+	//      keys
+	// ================================================
 
 	/// Deletes an API access key
 	pub async fn delete_key(&self, key: &mut Key) -> Result<(), Error> {
 		let state = key.state.as_option_mut().unwrap();
 
 		// --- done checking, now commit ---
-		// (the step at unset_local_bucket_alias will fail if a bucket
-		// does not have another alias, the deletion will be
-		// interrupted in the middle if that happens)
 
 		// 1. Delete local aliases
 		for (alias, _, to) in state.local_aliases.items().iter() {
 			if let Some(bucket_id) = to {
-				self.unset_local_bucket_alias(*bucket_id, &key.key_id, alias)
+				self.purge_local_bucket_alias(*bucket_id, &key.key_id, alias)
 					.await?;
 			}
 		}
@@ -415,4 +462,193 @@ impl<'a> LockedHelper<'a> {
 
 		Ok(())
 	}
+
+	// ================================================
+	//      repair procedure
+	// ================================================
+
+	pub async fn repair_aliases(&self) -> Result<(), GarageError> {
+		self.0.db.transaction(|tx| {
+			info!("--- begin repair_aliases transaction ----");
+
+			// 1. List all non-deleted buckets, so that we can fix bad aliases
+			let mut all_buckets: HashSet<Uuid> = HashSet::new();
+
+			for item in tx.range::<&[u8], _>(&self.0.bucket_table.data.store, ..)? {
+				let bucket = self
+					.0
+					.bucket_table
+					.data
+					.decode_entry(&(item?.1))
+					.map_err(db::TxError::Abort)?;
+				if !bucket.is_deleted() {
+					all_buckets.insert(bucket.id);
+				}
+			}
+
+			info!("number of buckets: {}", all_buckets.len());
+
+			// 2. List all aliases declared in bucket_alias_table and key_table
+			//    Take note of aliases that point to non-existing buckets
+			let mut global_aliases: HashMap<String, Uuid> = HashMap::new();
+
+			{
+				let mut delete_global = vec![];
+				for item in tx.range::<&[u8], _>(&self.0.bucket_alias_table.data.store, ..)? {
+					let mut alias = self
+						.0
+						.bucket_alias_table
+						.data
+						.decode_entry(&(item?.1))
+						.map_err(db::TxError::Abort)?;
+					if let Some(id) = alias.state.get() {
+						if all_buckets.contains(id) {
+							// keep aliases
+							global_aliases.insert(alias.name().to_string(), *id);
+						} else {
+							// delete alias
+							warn!(
+								"global alias: remove {} -> {:?} (bucket is deleted)",
+								alias.name(),
+								id
+							);
+							alias.state.update(None);
+							delete_global.push(alias);
+						}
+					}
+				}
+
+				info!("number of global aliases: {}", global_aliases.len());
+
+				info!("global alias table: {} entries fixed", delete_global.len());
+				for ga in delete_global {
+					debug!("Enqueue update to global alias table: {:?}", ga);
+					self.0.bucket_alias_table.queue_insert(tx, &ga)?;
+				}
+			}
+
+			let mut local_aliases: HashMap<(String, String), Uuid> = HashMap::new();
+
+			{
+				let mut delete_local = vec![];
+
+				for item in tx.range::<&[u8], _>(&self.0.key_table.data.store, ..)? {
+					let mut key = self
+						.0
+						.key_table
+						.data
+						.decode_entry(&(item?.1))
+						.map_err(db::TxError::Abort)?;
+					let Some(p) = key.state.as_option_mut() else {
+						continue;
+					};
+					let mut has_changes = false;
+					for (name, _, to) in p.local_aliases.items().to_vec() {
+						if let Some(id) = to {
+							if all_buckets.contains(&id) {
+								local_aliases.insert((key.key_id.clone(), name), id);
+							} else {
+								warn!(
+									"local alias: remove ({}, {}) -> {:?} (bucket is deleted)",
+									key.key_id, name, id
+								);
+								p.local_aliases.update_in_place(name, None);
+								has_changes = true;
+							}
+						}
+					}
+					if has_changes {
+						delete_local.push(key);
+					}
+				}
+
+				info!("number of local aliases: {}", local_aliases.len());
+
+				info!("key table: {} entries fixed", delete_local.len());
+				for la in delete_local {
+					debug!("Enqueue update to key table: {:?}", la);
+					self.0.key_table.queue_insert(tx, &la)?;
+				}
+			}
+
+			// 4. Reverse the alias maps to determine the aliases per-bucket
+			let mut bucket_global: HashMap<Uuid, Vec<String>> = HashMap::new();
+			let mut bucket_local: HashMap<Uuid, Vec<(String, String)>> = HashMap::new();
+
+			for (name, bucket) in global_aliases {
+				bucket_global.entry(bucket).or_default().push(name);
+			}
+			for ((key, name), bucket) in local_aliases {
+				bucket_local.entry(bucket).or_default().push((key, name));
+			}
+
+			// 5. Fix the bucket table to ensure consistency
+			let mut bucket_updates = vec![];
+
+			for item in tx.range::<&[u8], _>(&self.0.bucket_table.data.store, ..)? {
+				let bucket = self
+					.0
+					.bucket_table
+					.data
+					.decode_entry(&(item?.1))
+					.map_err(db::TxError::Abort)?;
+				let mut bucket2 = bucket.clone();
+				let Some(param) = bucket2.state.as_option_mut() else {
+					continue;
+				};
+
+				// fix global aliases
+				{
+					let ga = bucket_global.remove(&bucket.id).unwrap_or_default();
+					for (name, _, active) in param.aliases.items().to_vec() {
+						if active && !ga.contains(&name) {
+							warn!("bucket {:?}: remove global alias {}", bucket.id, name);
+							param.aliases.update_in_place(name, false);
+						}
+					}
+					for name in ga {
+						if param.aliases.get(&name).copied() != Some(true) {
+							warn!("bucket {:?}: add global alias {}", bucket.id, name);
+							param.aliases.update_in_place(name, true);
+						}
+					}
+				}
+
+				// fix local aliases
+				{
+					let la = bucket_local.remove(&bucket.id).unwrap_or_default();
+					for (pair, _, active) in param.local_aliases.items().to_vec() {
+						if active && !la.contains(&pair) {
+							warn!("bucket {:?}: remove local alias {:?}", bucket.id, pair);
+							param.local_aliases.update_in_place(pair, false);
+						}
+					}
+					for pair in la {
+						if param.local_aliases.get(&pair).copied() != Some(true) {
+							warn!("bucket {:?}: add local alias {:?}", bucket.id, pair);
+							param.local_aliases.update_in_place(pair, true);
+						}
+					}
+				}
+
+				if bucket2 != bucket {
+					bucket_updates.push(bucket2);
+				}
+			}
+
+			info!("bucket table: {} entries fixed", bucket_updates.len());
+			for b in bucket_updates {
+				debug!("Enqueue update to bucket table: {:?}", b);
+				self.0.bucket_table.queue_insert(tx, &b)?;
+			}
+
+			info!("--- end repair_aliases transaction ----");
+
+			Ok(())
+		})?;
+
+		info!("repair_aliases is done");
+
+		Ok(())
+	}
 }

src/model/s3/object_table.rs

@@ -259,7 +259,9 @@ mod v010 {
 			compressed: bool,
 			/// Whether the encryption uses an Object Encryption Key derived
 			/// from the master SSE-C key, instead of the master SSE-C key itself.
-			/// This is the case of objects created in Garage v2+
+			/// This is the case of objects created in Garage v2+.
+			/// This field is kept for compatibility with Garage v2.0.0-beta1,
+			/// which did not yet implement the v2 module below.
 			#[serde(default)]
 			use_oek: bool,
 		},
@@ -378,7 +380,222 @@ mod v010 {
 	}
 }
 
-pub use v010::*;
+mod v2 {
+	use garage_util::data::{Hash, Uuid};
+	use garage_util::migrate::Migrate;
+	use serde::{Deserialize, Serialize};
+
+	use super::v010;
+	pub use v010::{ChecksumAlgorithm, ChecksumValue};
+
+	/// An object
+	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	pub struct Object {
+		/// The bucket in which the object is stored, used as partition key
+		pub bucket_id: Uuid,
+
+		/// The key at which the object is stored in its bucket, used as sorting key
+		pub key: String,
+
+		/// The list of currently stored versions of the object
+		pub(super) versions: Vec<ObjectVersion>,
+	}
+
+	/// Information about a version of an object
+	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	pub struct ObjectVersion {
+		/// Id of the version
+		pub uuid: Uuid,
+		/// Timestamp of when the object was created
+		pub timestamp: u64,
+		/// State of the version
+		pub state: ObjectVersionState,
+	}
+
+	/// State of an object version
+	#[derive(PartialEq, Eq, Clone, Debug, Serialize, Deserialize)]
+	pub enum ObjectVersionState {
+		/// The version is being received
+		Uploading {
+			/// Indicates whether this is a multipart upload
+			multipart: bool,
+			/// Checksum algorithm and algorithm type to use
+			checksum_algorithm: Option<(ChecksumAlgorithm, ChecksumType)>,
+			/// Encryption params + headers to be included in the final object
+			encryption: ObjectVersionEncryption,
+		},
+		/// The version is fully received
+		Complete(ObjectVersionData),
+		/// The version uploaded containded errors or the upload was explicitly aborted
+		Aborted,
+	}
+
+	/// Data stored in object version
+	#[derive(PartialEq, Eq, PartialOrd, Ord, Clone, Debug, Serialize, Deserialize)]
+	pub enum ObjectVersionData {
+		/// The object was deleted, this Version is a tombstone to mark it as such
+		DeleteMarker,
+		/// The object is short, it's stored inlined.
+		/// It is never compressed. For encrypted objects, it is encrypted using
+		/// AES256-GCM, like the encrypted headers.
+		Inline(ObjectVersionMeta, #[serde(with = "serde_bytes")] Vec<u8>),
+		/// The object is not short, Hash of first block is stored here, next segments hashes are
+		/// stored in the version table
+		FirstBlock(ObjectVersionMeta, Hash),
+	}
+
+	/// Metadata about the object version
+	#[derive(PartialEq, Eq, PartialOrd, Ord, Clone, Debug, Serialize, Deserialize)]
+	pub struct ObjectVersionMeta {
+		/// Size of the object. If object is encrypted/compressed,
+		/// this is always the size of the unencrypted/uncompressed data
+		pub size: u64,
+		/// etag of the object
+		pub etag: String,
+		/// Encryption params + headers (encrypted or plaintext)
+		pub encryption: ObjectVersionEncryption,
+	}
+
+	/// Encryption information + metadata
+	#[derive(PartialEq, Eq, PartialOrd, Ord, Clone, Debug, Serialize, Deserialize)]
+	pub enum ObjectVersionEncryption {
+		SseC {
+			/// Encrypted serialized ObjectVersionInner struct.
+			/// This is never compressed, just encrypted using AES256-GCM.
+			#[serde(with = "serde_bytes")]
+			inner: Vec<u8>,
+			/// Whether data blocks are compressed in addition to being encrypted
+			/// (compression happens before encryption, whereas for non-encrypted
+			/// objects, compression is handled at the level of the block manager)
+			compressed: bool,
+			/// Whether the encryption uses an Object Encryption Key derived
+			/// from the master SSE-C key, instead of the master SSE-C key itself.
+			/// This is the case of objects created in Garage v2+
+			use_oek: bool,
+		},
+		Plaintext {
+			/// Plain-text headers
+			inner: ObjectVersionMetaInner,
+		},
+	}
+
+	/// Vector of headers, as tuples of the format (header name, header value)
+	/// Note: checksum can be Some(_) with checksum_type = None for objects that
+	/// have been migrated from Garage version before v2.0, as the distinction between
+	/// full-object and composite checksums was not implemented yet.
+	#[derive(PartialEq, Eq, PartialOrd, Ord, Clone, Debug, Serialize, Deserialize)]
+	pub struct ObjectVersionMetaInner {
+		pub headers: HeaderList,
+		pub checksum: Option<ChecksumValue>,
+		// checksum_type has to be stored separately, because when migrating
+		// from older versions of Garage, we can't know the correct value in
+		// ObjectVersionMetaInner::migrate (because it cannot take an argument
+		// that says whether the object was multipart or not)
+		pub checksum_type: Option<ChecksumType>,
+	}
+
+	pub type HeaderList = Vec<(String, String)>;
+
+	#[derive(PartialEq, Eq, PartialOrd, Ord, Clone, Copy, Debug, Serialize, Deserialize)]
+	pub enum ChecksumType {
+		FullObject,
+		Composite,
+	}
+
+	impl garage_util::migrate::Migrate for Object {
+		const VERSION_MARKER: &'static [u8] = b"G2s3ob";
+
+		type Previous = v010::Object;
+
+		fn migrate(old: v010::Object) -> Object {
+			Object {
+				bucket_id: old.bucket_id,
+				key: old.key,
+				versions: old.versions.into_iter().map(migrate_version).collect(),
+			}
+		}
+	}
+
+	fn migrate_version(old: v010::ObjectVersion) -> ObjectVersion {
+		ObjectVersion {
+			uuid: old.uuid,
+			timestamp: old.timestamp,
+			state: match old.state {
+				v010::ObjectVersionState::Uploading {
+					multipart,
+					checksum_algorithm,
+					encryption,
+				} => ObjectVersionState::Uploading {
+					multipart,
+					checksum_algorithm: checksum_algorithm.map(|algo| match multipart {
+						false => (algo, ChecksumType::FullObject),
+						true => (algo, ChecksumType::Composite),
+					}),
+					encryption: migrate_encryption(encryption),
+				},
+				v010::ObjectVersionState::Complete(d) => {
+					ObjectVersionState::Complete(migrate_data(d))
+				}
+				v010::ObjectVersionState::Aborted => ObjectVersionState::Aborted,
+			},
+		}
+	}
+
+	fn migrate_data(old: v010::ObjectVersionData) -> ObjectVersionData {
+		match old {
+			v010::ObjectVersionData::DeleteMarker => ObjectVersionData::DeleteMarker,
+			v010::ObjectVersionData::Inline(meta, data) => {
+				ObjectVersionData::Inline(migrate_meta(meta), data)
+			}
+			v010::ObjectVersionData::FirstBlock(meta, fb) => {
+				ObjectVersionData::FirstBlock(migrate_meta(meta), fb)
+			}
+		}
+	}
+
+	fn migrate_meta(old: v010::ObjectVersionMeta) -> ObjectVersionMeta {
+		ObjectVersionMeta {
+			size: old.size,
+			etag: old.etag,
+			encryption: migrate_encryption(old.encryption),
+		}
+	}
+
+	fn migrate_encryption(old: v010::ObjectVersionEncryption) -> ObjectVersionEncryption {
+		match old {
+			v010::ObjectVersionEncryption::SseC {
+				inner,
+				compressed,
+				use_oek,
+			} => ObjectVersionEncryption::SseC {
+				inner,
+				compressed,
+				use_oek,
+			},
+			v010::ObjectVersionEncryption::Plaintext { inner } => {
+				ObjectVersionEncryption::Plaintext {
+					inner: ObjectVersionMetaInner::migrate(inner),
+				}
+			}
+		}
+	}
+
+	impl Migrate for ObjectVersionMetaInner {
+		const VERSION_MARKER: &'static [u8] = b"G2s3om";
+
+		type Previous = v010::ObjectVersionMetaInner;
+
+		fn migrate(old: v010::ObjectVersionMetaInner) -> ObjectVersionMetaInner {
+			ObjectVersionMetaInner {
+				headers: old.headers,
+				checksum: old.checksum,
+				checksum_type: None,
+			}
+		}
+	}
+}
+
+pub use v2::*;
 
 impl Object {
 	/// Initialize an Object struct from parts

src/rpc/replication_mode.rs

@@ -71,26 +71,18 @@ pub fn parse_replication_mode(
 	config: &Config,
 ) -> Result<(ReplicationFactor, ConsistencyMode), Error> {
 	match (&config.replication_mode, config.replication_factor, config.consistency_mode.as_str()) {
-        (Some(replication_mode), None, "consistent") => {
-            tracing::warn!("Legacy config option replication_mode in use. Please migrate to replication_factor and consistency_mode");
-            let parsed_replication_mode = match replication_mode.as_str() {
-                "1" | "none"  => Some((ReplicationFactor(1), ConsistencyMode::Consistent)),
-                "2"           => Some((ReplicationFactor(2), ConsistencyMode::Consistent)),
-                "2-dangerous" => Some((ReplicationFactor(2), ConsistencyMode::Dangerous)),
-                "3"           => Some((ReplicationFactor(3), ConsistencyMode::Consistent)),
-                "3-degraded"  => Some((ReplicationFactor(3), ConsistencyMode::Degraded)),
-                "3-dangerous" => Some((ReplicationFactor(3), ConsistencyMode::Dangerous)),
-                _ => None,
-            };
-            Some(parsed_replication_mode.ok_or_message("Invalid replication_mode in config file.")?)
-        },
-        (None, Some(replication_factor), consistency_mode) => {
-            let replication_factor = ReplicationFactor::new(replication_factor)
-                .ok_or_message("Invalid replication_factor in config file.")?;
-            let consistency_mode = ConsistencyMode::parse(consistency_mode)
-                .ok_or_message("Invalid consistency_mode in config file.")?;
-            Some((replication_factor, consistency_mode))
-        }
-        _ => None,
-    }.ok_or_message("Either the legacy replication_mode or replication_level and consistency_mode can be set, not both.")
+		(Some(_replication_mode), _, _) => {
+			Err(Error::Message("The legacy replication_mode is no longer supported. Use replication_factor and consistency_mode instead.".into()))
+		}
+		(None, Some(replication_factor), consistency_mode) => {
+			let replication_factor = ReplicationFactor::new(replication_factor)
+				.ok_or_message("Invalid replication_factor in config file.")?;
+			let consistency_mode = ConsistencyMode::parse(consistency_mode)
+				.ok_or_message("Invalid consistency_mode in config file.")?;
+			Ok((replication_factor, consistency_mode))
+		}
+		(None, None, _) => {
+			Err(Error::Message("The option replication_factor is required.".into()))
+		}
+	}
 }

src/table/sync.rs

@@ -503,7 +503,7 @@ impl<F: TableSchema, R: TableReplication> SyncWorker<F, R> {
 
 	fn add_full_sync(&mut self) {
 		let mut partitions = self.syncer.data.replication.sync_partitions();
-		info!(
+		debug!(
 			"{}: Adding full sync for ack layout version {}",
 			F::TABLE_NAME,
 			partitions.layout_version
@@ -553,7 +553,7 @@ impl<F: TableSchema, R: TableReplication> Worker for SyncWorker<F, R> {
 			}
 
 			if todo.partitions.is_empty() {
-				info!(
+				debug!(
 					"{}: Completed full sync for ack layout version {}",
 					F::TABLE_NAME,
 					todo.layout_version

src/util/config.rs

@@ -135,6 +135,10 @@ pub struct Config {
 	/// Configuration for the admin API endpoint
 	#[serde(default = "Default::default")]
 	pub admin: AdminConfig,
+
+	/// Allow punycode in bucket names
+	#[serde(default)]
+	pub allow_punycode: bool,
 }
 
 /// Value for data_dir: either a single directory or a list of dirs with attributes

src/web/web_server.rs

@@ -469,10 +469,30 @@ fn error_to_res(e: Error) -> Response<BoxBody<Error>> {
 	//   was a HEAD request or we couldn't get the error document)
 	// We do NOT enter this code path when returning the bucket's
 	// error document (this is handled in serve_file)
-	let body = string_body(format!("{}\n", e));
-	let mut http_error = Response::new(body);
+	let mut body_str = format!(
+		r"<title>{http_code} {code_text}</title>
+<h1>{http_code} {code_text}</h1>",
+		http_code = e.http_status_code().as_u16(),
+		code_text = e.http_status_code().canonical_reason().unwrap_or("Unknown"),
+	);
+	if let Error::ApiError(ref err) = e {
+		body_str.push_str(&format!(
+			r"
+<ul>
+<li>Code: {s3_code}</li>
+<li>Message: {s3_message}.</li>
+</ul>",
+			s3_code = err.aws_code(),
+			s3_message = err,
+		));
+	}
+	let mut http_error = Response::new(string_body(body_str));
 	*http_error.status_mut() = e.http_status_code();
 	e.add_headers(http_error.headers_mut());
+	http_error.headers_mut().insert(
+		http::header::CONTENT_TYPE,
+		"text/html; charset=utf-8".parse().unwrap(),
+	);
 	http_error
 }