Files
orca/cloud/dev/scripts/relay-rehome-aggregate-evidence.test.mjs
Jinwoo Hong 68b11282a5 fix(relay): let the rehome evidence parser read a line the director grew (#21823)
The enable workflow reads the director's `[orca-relay] regional rehome
inventory` line out of Cloud Logging and pins the whole line with one regex.
Adding `hostNotArrivedLast24Hours` in #21813 made every healthy line stop
matching, so "Read fresh aggregate completion and abort evidence" threw
"no aggregate regional rehome inventory evidence" and the fail-closed step
disabled the durable switch at control generation 26.

The parser now requires the six original fields and tolerates further ones
in any order. Extra fields stay fenced by value shape rather than by pinning
the whole line: a field must be a bare name and a non-negative integer or
`none`, so `hostId=someone` is still not a counter and cannot ride along.
An absent count reads as null, not zero, because an older director not
reporting leaks is not the same as reporting none.

`hostNotArrivedLast24Hours` and `oldestActiveAgeMs` now reach the evidence
JSON and the operator step summary.

Two guards close the chain, each verified to fail on the regression it
exists for: a census in the relay package feeds the real formatter's output
to the real parser, and a script-side test pins the parser's output to the
fields the workflow summary renders.

Claude-Session: https://claude.ai/session/ced32ebb-7155-4413-adad-1eccd14c2010
2026-09-20 15:50:43 -04:00

136 lines
5.6 KiB
JavaScript

import assert from 'node:assert/strict'
import { test } from 'node:test'
import { parseRegionalRehomeInventory } from './relay-rehome-aggregate-evidence.mjs'
import { readRelayWorkflow } from './relay-repository.mjs'
const now = Date.parse('2026-08-14T12:00:00Z')
test('selects the newest fresh aggregate-only regional rehome inventory', () => {
const result = parseRegionalRehomeInventory([
{
timestamp: '2026-08-14T11:58:00Z',
textPayload: '[orca-relay] regional rehome inventory active=2 awaitingReceipt=1 targetRegistered=1 completedLast24Hours=9 abortedLast24Hours=0 oldestActiveAgeMs=30000'
},
{
timestamp: '2026-08-14T11:50:00Z',
textPayload: '[orca-relay] regional rehome inventory active=1 awaitingReceipt=0 targetRegistered=1 completedLast24Hours=8 abortedLast24Hours=0 oldestActiveAgeMs=none'
}
], { now, maxAgeMs: 5 * 60_000 })
assert.deepEqual(result, {
timestamp: Date.parse('2026-08-14T11:58:00Z'),
active: 2,
awaitingReceipt: 1,
targetRegistered: 1,
completedLast24Hours: 9,
abortedLast24Hours: 0,
hostNotArrivedLast24Hours: null,
oldestActiveAgeMs: 30_000
})
})
test('reads a line the director grew a field on, wherever the field sits', () => {
const result = parseRegionalRehomeInventory([{
timestamp: '2026-08-14T11:58:00Z',
textPayload:
'[orca-relay] regional rehome inventory hostNotArrivedLast24Hours=4 active=2' +
' awaitingReceipt=1 targetRegistered=1 completedLast24Hours=9 abortedLast24Hours=7' +
' oldestActiveAgeMs=30000 someFieldFromALaterRelease=11'
}], { now, maxAgeMs: 5 * 60_000 })
assert.equal(result.hostNotArrivedLast24Hours, 4)
assert.equal(result.abortedLast24Hours, 7)
assert.equal(result.oldestActiveAgeMs, 30_000)
})
test('reports an unmeasured host-not-arrived count as absent, not as zero', () => {
const [withField, withoutField] = ['4', null].map((value) =>
parseRegionalRehomeInventory([{
timestamp: '2026-08-14T11:58:00Z',
textPayload:
'[orca-relay] regional rehome inventory active=0 awaitingReceipt=0 targetRegistered=0' +
' completedLast24Hours=0 abortedLast24Hours=0 oldestActiveAgeMs=none' +
(value === null ? '' : ` hostNotArrivedLast24Hours=${value}`)
}], { now, maxAgeMs: 5 * 60_000 })
)
assert.equal(withField.hostNotArrivedLast24Hours, 4)
assert.equal(withoutField.hostNotArrivedLast24Hours, null)
})
test('rejects stale, malformed, and identity-bearing lookalikes', () => {
assert.throws(() => parseRegionalRehomeInventory([{
timestamp: '2026-08-14T11:00:00Z',
textPayload: '[orca-relay] regional rehome inventory active=0 awaitingReceipt=0 targetRegistered=0 completedLast24Hours=0 abortedLast24Hours=0 oldestActiveAgeMs=none'
}], { now, maxAgeMs: 5 * 60_000 }), /stale/)
assert.throws(() => parseRegionalRehomeInventory([{
timestamp: '2026-08-14T11:59:00Z',
textPayload: '[orca-relay] regional rehome inventory active=0 hostId=secret'
}], { now }), /no aggregate/)
})
test('keeps out an identity-bearing field riding along on a complete line', () => {
const complete =
'[orca-relay] regional rehome inventory active=0 awaitingReceipt=0 targetRegistered=0' +
' completedLast24Hours=0 abortedLast24Hours=0 oldestActiveAgeMs=none'
for (const extra of [' hostId=secret', ' userId=someone@example.test', ' note=a b']) {
assert.throws(
() => parseRegionalRehomeInventory(
[{ timestamp: '2026-08-14T11:59:00Z', textPayload: complete + extra }],
{ now }
),
/no aggregate/,
extra
)
}
})
test('refuses a line missing a required field, or repeating one', () => {
const missing =
'[orca-relay] regional rehome inventory active=0 awaitingReceipt=0 targetRegistered=0' +
' completedLast24Hours=0 oldestActiveAgeMs=none'
assert.throws(
() => parseRegionalRehomeInventory(
[{ timestamp: '2026-08-14T11:59:00Z', textPayload: missing }],
{ now }
),
/no aggregate/
)
assert.throws(
() => parseRegionalRehomeInventory(
[{ timestamp: '2026-08-14T11:59:00Z', textPayload: `${missing} abortedLast24Hours=0 abortedLast24Hours=1` }],
{ now }
),
/no aggregate/
)
assert.throws(
() => parseRegionalRehomeInventory(
[{
timestamp: '2026-08-14T11:59:00Z',
textPayload: missing.replace('active=0', 'active=none') + ' abortedLast24Hours=0'
}],
{ now }
),
/no aggregate/
)
})
// The third edge of the chain the enable workflow depends on. The formatter is
// pinned against this parser in the relay package's inventory-line census; this
// pins the parser against the summary an operator reads, so a field that
// reaches the evidence JSON and stops there fails here.
test('publishes every parsed counter in the operator step summary', () => {
const job = readRelayWorkflow('operate-relay-production-rehome-job.yml')
// The jq program and the file it reads sit on separate continuation lines, so
// match the whole render rather than one line of it.
const summary = /jq -r '([^']*)' \\\n\s*"\$\{RUNNER_TEMP\}\/relay-rehome-inventory\.json"/.exec(job)?.[1]
assert.ok(summary, 'the rehome job no longer renders the inventory evidence')
const evidence = parseRegionalRehomeInventory([{
timestamp: '2026-08-14T11:58:00Z',
textPayload:
'[orca-relay] regional rehome inventory active=0 awaitingReceipt=0 targetRegistered=0' +
' completedLast24Hours=0 abortedLast24Hours=0 hostNotArrivedLast24Hours=0 oldestActiveAgeMs=none'
}], { now, maxAgeMs: 5 * 60_000 })
for (const key of Object.keys(evidence)) {
if (key === 'timestamp') continue
assert.ok(summary.includes(`.${key}`), `${key} is missing from the step summary`)
}
})