Integrate reviewed affected-version troubleshooting help

Change-source: pulse-maintainer
This commit is contained in:
pulse-triage[bot]
2026-10-14 23:21:06 +01:00
3 changed files with 176 additions and 4 deletions
+46 -2
View File
@@ -11,6 +11,9 @@ first.** Do not run Diagnostics, force guest-agent probes or repeat a backup to
fill a report. A successful backup task or responsive console alone does not
establish safe recovery.
- **Not sure which version is affected:** distinguish the
[running server, reported agent and update target](TROUBLESHOOTING.md#find-the-running-server-and-agent-versions).
A helper's selected release or an agent's target is not its running version.
- **Cannot sign in:** use [local password recovery](TROUBLESHOOTING.md#i-forgot-my-password)
or [401 and login failures](TROUBLESHOOTING.md#cannot-login--401-unauthorized),
according to the existing authentication method. Do not delete configuration
@@ -72,6 +75,46 @@ valid, and a full configuration, journal or network export is not required.
## ⚡ Quick Fixes
### Find the running server and agent versions
Use the existing display on the affected Pulse installation; no update, restart
or new diagnostic run is needed just to identify a version.
- **Pulse server:** open **Settings → System → Updates** and read **Server
version**, not the available update. Keep any beta/RC suffix or development
build identity exactly as displayed. A newer server does not establish that
a separate agent has updated.
- **Pulse Agent:** under **Settings → Infrastructure**, open the affected
connection and read **Pulse Agent version** alongside **Last seen**. An
**Update available** value such as `6.4.5 → 6.5.0` distinguishes the reported
version on the left from the target on the right. **Version target** alone
is not a reported agent version. If the connection is stale or offline, its
last report does not prove what is running now; keep that time and uncertainty.
For an API-only connection with no Pulse Agent, report the agent as **none**,
not the server's version. See [agent update behaviour](UNIFIED_AGENT.md#auto-update)
for the separate update path.
- **Container image:** include the image tag or digest already available in
your deployment manager, separately from **Server version**. A tag such as
`latest` is mutable; a Compose file or selected image is not proof that a
running container was replaced. Keep the exact reference, including the
registry and preview suffix; do not substitute a GitHub release tag. A
`helm-chart-…` release identifies a chart, not a Pulse server archive or the
application version running in Kubernetes.
Keep the **version at the failure**, an interrupted update's **intended version**
and the **version observed after recovery** separate. Today's display does not
reconstruct the failing run. If the helper selected a release but Pulse never
started, give that attempted release and the public helper name or URL instead;
installer output or a verified download signature does not prove completion.
If a view is unavailable or you cannot access it with your existing permissions,
write **unknown**; do not widen access, reinstall or re-enrol an agent, rerun an
update or enable Debug to fill the gap. Do not share full container inspections,
resolved Compose configuration, service environments or command lines: they can
contain credentials. Existing version text or a locally reviewed, redacted
excerpt is enough. Add it to the [existing report](#-getting-help); information
already supplied does not need to be collected again.
### I forgot my password
Pulse does not provide an email password-reset flow. Choose the path that
@@ -1882,8 +1925,9 @@ If you're still stuck:
exact error. Do not repeat an update, outage or notification storm merely to
reproduce it. A failed update banner does not prove the action left the
target unchanged; check its current state before another attempt.
2. **Identify the affected version**: give the running Pulse and relevant agent
versions, not just the version before an upgrade. For Docker, include the
2. **Identify the affected version**: distinguish the
[running server, reported agent and update target](#find-the-running-server-and-agent-versions),
not just the version before an upgrade. For Docker, include the
running image tag or digest. If installation never started Pulse, give the
attempted release and public installer/helper source, or say "unknown".
3. **Choose relevant, safe evidence**: use existing logs, errors or observations
+46 -2
View File
@@ -11,6 +11,9 @@ first.** Do not run Diagnostics, force guest-agent probes or repeat a backup to
fill a report. A successful backup task or responsive console alone does not
establish safe recovery.
- **Not sure which version is affected:** distinguish the
[running server, reported agent and update target](TROUBLESHOOTING.md#find-the-running-server-and-agent-versions).
A helper's selected release or an agent's target is not its running version.
- **Cannot sign in:** use [local password recovery](TROUBLESHOOTING.md#i-forgot-my-password)
or [401 and login failures](TROUBLESHOOTING.md#cannot-login--401-unauthorized),
according to the existing authentication method. Do not delete configuration
@@ -72,6 +75,46 @@ valid, and a full configuration, journal or network export is not required.
## ⚡ Quick Fixes
### Find the running server and agent versions
Use the existing display on the affected Pulse installation; no update, restart
or new diagnostic run is needed just to identify a version.
- **Pulse server:** open **Settings → System → Updates** and read **Server
version**, not the available update. Keep any beta/RC suffix or development
build identity exactly as displayed. A newer server does not establish that
a separate agent has updated.
- **Pulse Agent:** under **Settings → Infrastructure**, open the affected
connection and read **Pulse Agent version** alongside **Last seen**. An
**Update available** value such as `6.4.5 → 6.5.0` distinguishes the reported
version on the left from the target on the right. **Version target** alone
is not a reported agent version. If the connection is stale or offline, its
last report does not prove what is running now; keep that time and uncertainty.
For an API-only connection with no Pulse Agent, report the agent as **none**,
not the server's version. See [agent update behaviour](UNIFIED_AGENT.md#auto-update)
for the separate update path.
- **Container image:** include the image tag or digest already available in
your deployment manager, separately from **Server version**. A tag such as
`latest` is mutable; a Compose file or selected image is not proof that a
running container was replaced. Keep the exact reference, including the
registry and preview suffix; do not substitute a GitHub release tag. A
`helm-chart-…` release identifies a chart, not a Pulse server archive or the
application version running in Kubernetes.
Keep the **version at the failure**, an interrupted update's **intended version**
and the **version observed after recovery** separate. Today's display does not
reconstruct the failing run. If the helper selected a release but Pulse never
started, give that attempted release and the public helper name or URL instead;
installer output or a verified download signature does not prove completion.
If a view is unavailable or you cannot access it with your existing permissions,
write **unknown**; do not widen access, reinstall or re-enrol an agent, rerun an
update or enable Debug to fill the gap. Do not share full container inspections,
resolved Compose configuration, service environments or command lines: they can
contain credentials. Existing version text or a locally reviewed, redacted
excerpt is enough. Add it to the [existing report](#-getting-help); information
already supplied does not need to be collected again.
### I forgot my password
Pulse does not provide an email password-reset flow. Choose the path that
@@ -1882,8 +1925,9 @@ If you're still stuck:
exact error. Do not repeat an update, outage or notification storm merely to
reproduce it. A failed update banner does not prove the action left the
target unchanged; check its current state before another attempt.
2. **Identify the affected version**: give the running Pulse and relevant agent
versions, not just the version before an upgrade. For Docker, include the
2. **Identify the affected version**: distinguish the
[running server, reported agent and update target](#find-the-running-server-and-agent-versions),
not just the version before an upgrade. For Docker, include the
running image tag or digest. If installation never started Pulse, give the
attempted release and public installer/helper source, or say "unknown".
3. **Choose relevant, safe evidence**: use existing logs, errors or observations
@@ -0,0 +1,84 @@
#!/usr/bin/env python3
"""Keep passive version help tied to the real current/report/target displays."""
from pathlib import Path
import unittest
ROOT = Path(__file__).resolve().parents[2]
HEADING = "### Find the running server and agent versions\n"
ANCHOR = "#find-the-running-server-and-agent-versions"
class RunningVersionHelpDocsTest(unittest.TestCase):
def setUp(self):
self.source = (ROOT / "docs/TROUBLESHOOTING.md").read_text()
self.guide = " ".join(
self.source.split(HEADING, 1)[1].split("\n### ", 1)[0].split()
)
def test_workflow_entry_and_getting_help_reach_the_same_guidance(self):
self.assertIn(ANCHOR, self.source.split("## ⚡ Quick Fixes", 1)[0])
self.assertIn(ANCHOR, self.source.split("## 🆘 Getting Help", 1)[1])
self.assertEqual(self.source.count(HEADING), 1)
def test_server_instruction_names_the_running_field_not_the_update(self):
copy = (ROOT / "frontend-modern/src/utils/updatesPresentation.ts").read_text()
panel = (ROOT / "frontend-modern/src/components/Settings/UpdatesSettingsPanel.tsx").read_text()
self.assertIn("currentVersionLabel: 'Server version'", copy)
self.assertIn("{props.versionInfo()?.version || 'Loading...'}", panel)
self.assertIn("**Server version**, not the available update", self.guide)
self.assertIn("Keep any beta/RC suffix or development build identity", self.guide)
self.assertIn("A newer server does not establish", self.guide)
def test_agent_instruction_distinguishes_last_report_from_target(self):
model = (ROOT / "frontend-modern/src/components/Settings/connectionsTableModel.ts").read_text()
workspace = (ROOT / "frontend-modern/src/components/Settings/InfrastructureWorkspace.tsx").read_text()
self.assertIn("connection.agentVersion?.trim()", model)
self.assertIn("connection.expectedAgentVersion?.trim()", model)
self.assertIn("detail: `${currentVersion} -> ${expectedVersion}`", model)
self.assertIn("badgeLabel: 'Version target'", model)
self.assertIn("Pulse Agent version", workspace)
self.assertIn("infoCard('Last seen'", workspace)
for fact in (
"reported version on the left", "target on the right",
"**Version target** alone is not a reported agent version",
"last report does not prove what is running now",
"agent as **none**, not the server's version",
"UNIFIED_AGENT.md#auto-update",
):
with self.subTest(fact=fact):
self.assertIn(fact, self.guide)
def test_image_helper_and_recovery_are_not_treated_as_running_source(self):
for fact in (
"`latest` is mutable", "not proof that a running container was replaced",
"do not substitute a GitHub release tag", "`helm-chart-…`",
"not a Pulse server archive", "**version at the failure**",
"**intended version**", "**version observed after recovery**",
"Today's display does not reconstruct the failing run",
"verified download signature does not prove completion",
):
with self.subTest(fact=fact):
self.assertIn(fact, self.guide)
def test_unknown_is_valid_without_mutation_or_private_exports(self):
self.assertNotIn("```", self.guide)
for boundary in (
"write **unknown**", "do not widen access, reinstall or re-enrol",
"rerun an update or enable Debug", "Do not share full container inspections",
"resolved Compose configuration", "service environments or command lines",
"locally reviewed, redacted excerpt", "already supplied does not need",
):
with self.subTest(boundary=boundary):
self.assertIn(boundary, self.guide)
def test_shipped_help_is_byte_identical(self):
self.assertEqual(
(ROOT / "docs/TROUBLESHOOTING.md").read_bytes(),
(ROOT / "frontend-modern/public/docs/TROUBLESHOOTING.md").read_bytes(),
)
if __name__ == "__main__":
unittest.main()