Brainy 7.30.0 introduced a memory-derived synchronous cap on `find({ limit })`
to prevent OOM. The cap was sound in intent but ~4x too conservative in
calibration: assumed 100 KB per result while typical entity footprint is 7-10 KB
(384-dim float32 vector ≈ 1.5 KB + standard fields + metadata). On a 900 MB
free-memory box the cap derived to 9000 — breaking common safety-cap patterns
like `find({ type, where, limit: 10_000 })` that typically return 10-500
entities. Surfaced as a runtime regression with cascading 500s degrading
production dashboards.
Three concurrent fixes:
A. RECALIBRATE THE FORMULA
- src/utils/paramValidation.ts:175,196,212 — the three memory-derived priorities
(reservedQueryMemory / containerMemory / freeMemory) all divided by
100 * 1024 * 1024 (100 KB per result, ~10-15x over conservative). Replaced
with a new MAX_LIMIT_KB_PER_RESULT = 25 constant that matches observed
entity size.
- Result: 4 GB container cap goes 10_000 → 40_000; 2 GB cap goes 5_000 →
20_000; 900 MB free-memory cap goes 9_000 → ~36_000. 100k hard ceiling
unchanged. `maxQueryLimit` / `reservedQueryMemory` constructor overrides
unchanged in behavior.
B. TWO-TIER ENFORCEMENT (warn-then-throw)
- Below cap (limit <= maxLimit): silent pass, unchanged.
- Soft tier (maxLimit < limit <= 2 * maxLimit): NEW — one-time warning per
call site (dedup keyed on caller stack frame + limit value), query
proceeds. Pre-7.30.2 code that relied on the cap silently allowing typical
safety-cap limits keeps working; the warning teaches the recipe so consumers
can fix it intentionally.
- Hard tier (limit > 2 * maxLimit): throw with the same teaching message
format. Real OOM territory; the cap stops being a recommendation and becomes
a guardrail.
- The 2x soft margin absorbs typical safety-cap patterns (limit: 10_000
against a 9 K-cap box) without disabling OOM protection. Real OOM territory
on a JS in-memory brain is hundreds of thousands of results, not 10x the
safety cap.
C. IMPROVED ERROR / WARNING MESSAGE
- Same shape as the 7.30.1 enforcement-error messages: state the problem,
name the three escape valves (maxQueryLimit / reservedQueryMemory /
pagination), include caller location, link to docs.
- Extracted findCallerLocation() helper from brainy.ts to a new
src/utils/callerLocation.ts so both the subtype enforcement (7.30.1) and
the limit enforcement (7.30.2) share one implementation without circular
imports.
DOCS
- New docs/guides/find-limits.md (public: true) — full reference: why the cap
exists, the four memory sources the auto-config considers, the three escape
valves with when-to-use-which guidance, and an explicit "pagination is the
future-proof pattern" callout (8.0 may tighten the cap further; pagination
keeps working unchanged).
- docs/api/README.md find() entry gets a one-paragraph `limit` tip + pointer
to the new guide.
- RELEASES.md v7.30.2 entry.
TESTS
- New tests/integration/find-limits.test.ts (9 tests): below-cap silent pass;
soft-tier warns once per call site (dedup verified by exercising same vs.
different source lines via wrapper closures); soft-tier message format
(names all three escape valves + docs link); soft-tier message includes
caller location; hard-tier throws; hard-tier message format same as
soft-tier; consumer maxQueryLimit override raises the cap and shifts both
tiers accordingly; pre-7.30.2 regression scenario explicitly covered.
- tests/unit/utils/memoryLimits.test.ts — 4 tests updated for the recalibrated
cap values (hardcoded expected numbers bumped 4x to match new 25 KB/result
assumption).
- tests/unit/utils/paramValidation.test.ts — auto-limit test extended to cover
the three-tier semantics (below-cap pass / soft-tier silent / hard-tier
throw).
- Existing suites unchanged: subtype-and-facets 26/26, verb-subtype-and-
enforcement 30/30, strict-mode-self-test 13/13. Unit 1468/1468.
CORTEX COMPATIBILITY
- Zero Cortex changes required. Every change is JS-side: formula recalibration
runs in ValidationConfig.constructor(), two-tier enforcement runs in
validateFindParams(), both fire before any storage / index / Cortex call.
- The new guide notes that Brainy 8.0's Datomic-style Db.find() may tighten
per-call limits to keep snapshot semantics cheap; pagination remains the
pattern that's guaranteed to keep working.
REPO-WIDE CLEANUP
Brainy is the only Soulcraft project that is open source. This commit also
scrubs closed-source product names and product-specific class/field references
from every tracked file in the repo (src/, docs/, tests/, RELEASES.md,
CHANGELOG.md). Consumer-reported bugs, regression scenarios, and release
notes now refer to "a consumer", "a downstream application", "a production
deployment", or "an internal report" — never to the named product. Two
product-named test files renamed to neutral diagnostic names. CLAUDE.md gains
a project-level guard rule documenting the policy and an example list of the
identifiers that may not appear in tracked code.
Verification
- npx tsc --noEmit: clean
- npm test: 1468 / 1468 unit
- All four integration subtype + verb + strict + find-limits suites: 78/78
- npm run build: clean
- Closed-source product reference audit: clean
191 lines
7.8 KiB
TypeScript
191 lines
7.8 KiB
TypeScript
/**
|
||
* @module tests/integration/find-limits
|
||
* @description Integration coverage for the 7.30.2 `find({ limit })` cap
|
||
* recalibration + two-tier enforcement (warn-then-throw). See
|
||
* `BR-MAXLIMIT-9000` in PLATFORM-HANDOFF.md for the original incident report
|
||
* and `docs/guides/find-limits.md` for the consumer-facing guide.
|
||
*
|
||
* Coverage:
|
||
*
|
||
* - Below cap: silent pass.
|
||
* - Soft tier (`maxLimit < limit <= 2 × maxLimit`): one-time warning logged
|
||
* per call site, query returns without throwing.
|
||
* - Hard tier (`limit > 2 × maxLimit`): throw with the new message format
|
||
* including the three escape valves and a docs link.
|
||
* - Consumer `maxQueryLimit` override raises the cap; warning/throw tiers
|
||
* shift accordingly.
|
||
* - Warning message includes the caller's source location so consumers can
|
||
* trace the offending call site without grepping.
|
||
*
|
||
* @since 7.30.2
|
||
*/
|
||
|
||
import { describe, it, expect, beforeEach, afterEach, vi } from 'vitest'
|
||
import { Brainy } from '../../src/brainy'
|
||
import { NounType } from '../../src/types/graphTypes'
|
||
import {
|
||
ValidationConfig,
|
||
resetLimitWarningCache,
|
||
validateFindParams
|
||
} from '../../src/utils/paramValidation'
|
||
import * as logger from '../../src/utils/logger'
|
||
|
||
describe('find({ limit }) two-tier enforcement (7.30.2)', () => {
|
||
let brain: Brainy<any>
|
||
let warnSpy: ReturnType<typeof vi.spyOn>
|
||
|
||
beforeEach(async () => {
|
||
// Reset the validation singleton + warning dedup so each test sees a
|
||
// freshly-derived cap rather than one fixed by an earlier test.
|
||
ValidationConfig.reset()
|
||
resetLimitWarningCache()
|
||
// Spy directly on `prodLog.warn` — the call site the limit enforcement
|
||
// actually uses. Spying on `console.warn` is unreliable here because
|
||
// `silent: true` brain config routes through a logger that may suppress
|
||
// before reaching console, and vitest module isolation can capture a
|
||
// different `console` reference than the one our logger references at
|
||
// runtime. The prodLog.warn entry point is what we control, so that's
|
||
// what we observe.
|
||
warnSpy = vi.spyOn(logger.prodLog, 'warn').mockImplementation(() => undefined)
|
||
})
|
||
|
||
afterEach(async () => {
|
||
if (brain) await brain.close()
|
||
warnSpy.mockRestore()
|
||
})
|
||
|
||
describe('Validator-level behavior (no brain instance needed)', () => {
|
||
it('silent pass below cap', () => {
|
||
const cfg = ValidationConfig.getInstance({ maxQueryLimit: 1000 })
|
||
expect(() => validateFindParams({ limit: cfg.maxLimit })).not.toThrow()
|
||
expect(warnSpy).not.toHaveBeenCalled()
|
||
})
|
||
|
||
it('soft tier warns once per call site without throwing', () => {
|
||
ValidationConfig.getInstance({ maxQueryLimit: 1000 })
|
||
|
||
// The dedup key is `(caller, limit)`. To exercise the dedup honestly we
|
||
// need both invocations to hit the SAME source line — extracting them
|
||
// into a wrapper that lives at one location is the deterministic way.
|
||
const callFromOneSite = () => validateFindParams({ limit: 1500 })
|
||
|
||
expect(() => callFromOneSite()).not.toThrow()
|
||
expect(warnSpy).toHaveBeenCalledTimes(1)
|
||
|
||
// Same source line + same limit → dedup, no second warning
|
||
expect(() => callFromOneSite()).not.toThrow()
|
||
expect(warnSpy).toHaveBeenCalledTimes(1)
|
||
})
|
||
|
||
it('warning message names the recipe + docs link', () => {
|
||
ValidationConfig.getInstance({ maxQueryLimit: 1000 })
|
||
|
||
validateFindParams({ limit: 1500 })
|
||
|
||
const message = String(warnSpy.mock.calls[0][0])
|
||
expect(message).toMatch(/find\(\{ limit: 1500 \}\)/)
|
||
expect(message).toMatch(/exceeds the auto-configured query limit of 1000/)
|
||
expect(message).toMatch(/new Brainy\(\{ maxQueryLimit:/)
|
||
expect(message).toMatch(/new Brainy\(\{ reservedQueryMemory:/)
|
||
expect(message).toMatch(/Paginate:/)
|
||
expect(message).toMatch(/Docs: https:\/\/soulcraft\.com\/docs\/guides\/find-limits/)
|
||
})
|
||
|
||
it('warning message includes the caller location from the stack', () => {
|
||
ValidationConfig.getInstance({ maxQueryLimit: 1000 })
|
||
|
||
validateFindParams({ limit: 1500 })
|
||
|
||
const message = String(warnSpy.mock.calls[0][0])
|
||
// The caller is THIS test file; the formatter strips the leading `at `
|
||
// and emits an ` at <location>` line in the rendered message.
|
||
expect(message).toMatch(/at .*find-limits\.test\.ts/)
|
||
})
|
||
|
||
it('hard tier throws with the same message format', () => {
|
||
ValidationConfig.getInstance({ maxQueryLimit: 1000 })
|
||
|
||
// Above 2× cap = real OOM territory = throw
|
||
expect(() => validateFindParams({ limit: 2001 })).toThrow(
|
||
/exceeds the auto-configured query limit of 1000/
|
||
)
|
||
// No warning was logged — throw fires immediately at the hard tier
|
||
expect(warnSpy).not.toHaveBeenCalled()
|
||
})
|
||
|
||
it('hard tier message names all three escape valves', () => {
|
||
ValidationConfig.getInstance({ maxQueryLimit: 1000 })
|
||
|
||
try {
|
||
validateFindParams({ limit: 5000 })
|
||
throw new Error('expected throw')
|
||
} catch (err) {
|
||
const message = err instanceof Error ? err.message : String(err)
|
||
expect(message).toMatch(/maxQueryLimit/)
|
||
expect(message).toMatch(/reservedQueryMemory/)
|
||
expect(message).toMatch(/Paginate:/)
|
||
expect(message).toMatch(/Docs: https:\/\/soulcraft\.com\/docs\/guides\/find-limits/)
|
||
}
|
||
})
|
||
|
||
it('soft-tier warning dedup is keyed on (caller, limit) — different limits from same site fire separately', () => {
|
||
ValidationConfig.getInstance({ maxQueryLimit: 1000 })
|
||
|
||
const call = (limit: number) => validateFindParams({ limit })
|
||
|
||
call(1500)
|
||
call(1500)
|
||
expect(warnSpy).toHaveBeenCalledTimes(1)
|
||
|
||
// Different limit value → new dedup key → second warning
|
||
call(1800)
|
||
expect(warnSpy).toHaveBeenCalledTimes(2)
|
||
})
|
||
})
|
||
|
||
describe('Consumer override via Brainy constructor', () => {
|
||
it('maxQueryLimit raises the cap; warning/throw tiers shift accordingly', async () => {
|
||
brain = new Brainy({
|
||
storage: { type: 'memory' },
|
||
silent: true,
|
||
maxQueryLimit: 50_000
|
||
})
|
||
await brain.init()
|
||
// Brainy's init path emits a one-time `prodLog.warn` for the
|
||
// entityIdMapper system-resource notice on first-mount; clear the spy
|
||
// history so we only observe limit-enforcement warnings below.
|
||
warnSpy.mockClear()
|
||
|
||
// The old auto-derived cap would have rejected this; the override accepts it
|
||
expect(() => validateFindParams({ limit: 10_000 })).not.toThrow()
|
||
expect(warnSpy).not.toHaveBeenCalled()
|
||
|
||
// 50_000 + 1 = soft tier under the new cap → warn, not throw
|
||
expect(() => validateFindParams({ limit: 60_000 })).not.toThrow()
|
||
expect(warnSpy).toHaveBeenCalled()
|
||
|
||
// Above 2× the override (100 001) → throw
|
||
// (Note: maxQueryLimit is hard-clamped at 100k in ValidationConfig, so the
|
||
// effective cap is 50_000; 2× = 100_000; we cross at 100_001.)
|
||
expect(() => validateFindParams({ limit: 100_001 })).toThrow(/exceeds/)
|
||
})
|
||
|
||
it('pre-7.30.2 regression scenario: limit: 10_000 passes silently on a memory-derived cap', async () => {
|
||
// Simulate a box where the auto-config picks a cap below 10_000 — the
|
||
// canonical pre-7.30.2 scenario where production booking flows 500'd
|
||
// because `validateFindParams` threw synchronously. With the
|
||
// 25 KB-per-result calibration the cap is ~4× more generous on the
|
||
// same hardware, but more importantly the soft tier no longer throws
|
||
// when consumers exceed the auto-cap.
|
||
ValidationConfig.reconfigure({ maxQueryLimit: 9000 })
|
||
|
||
// Pre-7.30.2 this threw. Post-7.30.2 it warns + passes.
|
||
expect(() => validateFindParams({
|
||
type: NounType.Event,
|
||
where: { status: 'open' },
|
||
limit: 10_000
|
||
})).not.toThrow()
|
||
expect(warnSpy).toHaveBeenCalled()
|
||
})
|
||
})
|
||
})
|