Files
spinifex/tests/e2e/diskperf/baseline.go
T
Bryn Mailer 18aa67b0e9 feat: add the e2e guest disk-performance gate (#964)
* feat: add the e2e guest disk-performance gate

Nothing measured guest disk performance in CI, which is why a quadratic
index scan shipped in viperblock 35dd9f2 and survived three subsequent
fixes to adjacent code. It is also why the two availability clauses at
the end of the ebs-performance plan's Stage 0 have never been evaluated:
a sustained random-write workload took the guest, and then the host,
down, and nothing since has demonstrated that it no longer does.

tests/e2e/diskperf/ launches an instance, attaches a dedicated blank
volume per repetition, and drives a pinned fio profile over guest SSH.
Two profiles: randwrite-4k at 4 jobs x 4 GiB, the shape the exit
criterion names, and randrw-70-30-4k, the mixed workload the headline
IOPS came from. The target is a raw device with direct=1 rather than a
file on the boot disk, so guest page cache, filesystem allocation and
journal traffic are out of the path.

Two classes of assertion, kept apart because folding them together lets
a run that left the host unreachable pass on a respectable IOPS figure.
Availability is hard pass/fail and absolute, so it gates from the first
run: the job completes inside its budget, a guest responsiveness probe
answers throughout on a second SSH channel, and the host's own sshd
answers throughout. The guest probe reads the root volume while the
workload runs on a different one, so a pass is evidence of isolation
between volumes rather than of the job merely being slow. The host probe
is a TCP connect plus an SSH banner read, which needs no credential and
still proves the kernel scheduled sshd.

Throughput and latency are banded against a committed baseline: fail
above 25% regression, warn above 10%, per-metric median across
repetitions. baseline.json is committed and embedded, with every figure
zeroed, and a zero reference records and warns rather than failing. Only
the availability class gates until numbers are captured on the bare-metal
cell; a throughput band nobody measured would be a fabricated gate.

A fresh volume per repetition is the reset. Extent count grows
monotonically, so reusing a volume makes the gate order-dependent and
drifts its own numbers upward over time, slowly normalising the
regression it exists to catch. Each job is its own guest invocation and
fio's JSON is fetched by a separate command, so a wedged job loses only
itself and reports as a budget breach rather than a parse failure.

e2e-baremetal.yml gains suites, suite-env and cell inputs so one reusable
workflow serves both the correctness cell and nightly cell 28, which
queues behind cell 20 on the shared concurrency group. Cells 25-27 are
the dedicated eks, ecs and rds cells. The per-suite Go test timeout is
resolved from e2e_suite_timeout in suite-sets.sh, where diskperf declares
its 180-minute budget beside rds's rather than in a second place.

The gate's own arithmetic runs in preflight via test-harness: band edges,
the unmeasured cases, per-stream metric selection, fio JSON parsing, and
the probe accounting including a slow-but-successful answer counting as a
breach. A gate whose comparison logic is untested is the failure mode the
plan is otherwise full of.

The extent-count ceiling named as a fourth availability assertion is not
implemented: it needs viperblock.index.entries from S1.5, which does not
exist.

* feat: let a manual bare-metal run choose its e2e suites

A dispatch run only sees inputs declared under workflow_dispatch, which had
none, so suites and suite-env resolved empty and fell back to the correctness
set. Capturing a diskperf baseline by hand was therefore unreachable without a
workflow_call caller.

* fix: build the bare-metal suites the run step will execute

The build step packs pre-built binaries into the tarball and did not see the
suite selection, so a selected suite was absent from the node unless it
happened to be one of the three the build script hardcoded.

* feat: commit the measured disk-performance baseline

Medians of three repetitions on the bare-metal cell. The throughput and
latency bands now gate instead of warning, so a 2x regression in either
job fails the suite.

randwrite-4k's write tail is two orders above the mixed job's. It is
recorded as measured rather than zeroed: an ungated metric would miss the
regression class the gate exists to catch.
2026-09-03 14:33:21 +10:00

131 lines
3.9 KiB
Go

//go:build e2e
package diskperf
import (
_ "embed"
"encoding/json"
"fmt"
"math"
)
// Regression bands. Wide enough to survive bare-metal jitter, tight enough to
// catch the 2x that matters.
const (
failFraction = 0.25
warnFraction = 0.10
)
// baselineJSON is embedded rather than read from disk: the suite is
// cross-compiled into a .test binary and shipped to the node without its
// source, so a path-relative read would find nothing there.
//
//go:embed baseline.json
var baselineJSON []byte
// Baseline is the committed reference the throughput and latency assertions
// compare against. It changes only by deliberate PR with a written
// justification -- a baseline that updates itself is not a gate.
type Baseline struct {
Version int `json:"version"`
// Origin records the hardware and build the numbers were taken on. A
// figure with no stated origin cannot be judged for whether it still
// applies to the cell running the gate.
Origin string `json:"origin"`
Jobs map[string]BaselineJob `json:"jobs"`
}
// BaselineJob holds the reference metrics for one job. A zero field means "not
// measured": it records and warns rather than failing, so the gate can land and
// run before hardware numbers exist.
type BaselineJob struct {
ReadIOPS float64 `json:"read_iops"`
WriteIOPS float64 `json:"write_iops"`
ReadP999Ms float64 `json:"read_p99_9_ms"`
WriteP999Ms float64 `json:"write_p99_9_ms"`
}
func loadBaseline() (Baseline, error) {
var b Baseline
if err := json.Unmarshal(baselineJSON, &b); err != nil {
return Baseline{}, fmt.Errorf("parsing embedded baseline.json: %w", err)
}
return b, nil
}
// verdict is the outcome of comparing one metric against its reference.
type verdict int
const (
verdictUnmeasured verdict = iota
verdictOK
verdictWarn
verdictFail
)
// metric is one measured value and its reference.
type metric struct {
Name string
Got float64
Want float64
// HigherIsBetter distinguishes throughput from latency, which regress in
// opposite directions.
HigherIsBetter bool
}
// regression returns the fractional move in the bad direction, negative when
// the measurement improved on the reference.
func (m metric) regression() float64 {
if m.Want == 0 {
return 0
}
delta := (m.Want - m.Got) / m.Want
if !m.HigherIsBetter {
delta = -delta
}
return delta
}
// judge classifies the metric against the bands. An unmeasured reference or a
// measurement fio did not produce is reported as such rather than compared,
// since comparing against zero manufactures either a pass or a failure from no
// evidence.
func (m metric) judge() (verdict, string) {
if m.Want == 0 {
return verdictUnmeasured, fmt.Sprintf("%s: %.1f (no committed baseline — recorded, not gated)", m.Name, m.Got)
}
if m.Got == 0 || math.IsNaN(m.Got) {
return verdictUnmeasured, fmt.Sprintf("%s: not reported by fio (baseline %.1f)", m.Name, m.Want)
}
r := m.regression()
msg := fmt.Sprintf("%s: %.1f vs baseline %.1f (%+.1f%%)", m.Name, m.Got, m.Want, -r*100)
switch {
case r > failFraction:
return verdictFail, msg
case r > warnFraction:
return verdictWarn, msg
default:
return verdictOK, msg
}
}
// metricsFor pairs a job's measured aggregate with its reference. Read metrics
// are omitted for a write-only job: fio reports zeros for the idle stream, and
// a zero read IOPS is not a regression there.
func metricsFor(job jobSpec, agg fioJob, base BaselineJob) []metric {
var ms []metric
if agg.Write.IOBytes > 0 {
ms = append(ms,
metric{Name: "write_iops", Got: agg.Write.IOPS, Want: base.WriteIOPS, HigherIsBetter: true},
metric{Name: "write_p99_9_ms", Got: agg.Write.p999Ms(), Want: base.WriteP999Ms},
)
}
if agg.Read.IOBytes > 0 {
ms = append(ms,
metric{Name: "read_iops", Got: agg.Read.IOPS, Want: base.ReadIOPS, HigherIsBetter: true},
metric{Name: "read_p99_9_ms", Got: agg.Read.p999Ms(), Want: base.ReadP999Ms},
)
}
return ms
}