Commit Graph
12 Commits
Author SHA1 Message Date
ZyyandCodex 3e33e84a45 fix(observability): unify request error statistics
Count incomplete requests as errors, exclude recovered and cancelled requests from success rates, and preserve gaps for missing samples.

Co-authored-by: Codex <noreply@openai.com>
2026-11-05 21:56:12 +08:00
3a2ccbc5e6 perf(usage): query insights by card and selected metric (#556)
* fix(usage): keep optional concurrency scans within a bounded budget

Skip unused concurrency work for Key usage and Dashboard trends. Preserve
other insights when exact concurrency exceeds its row, time, or slot budget.

Constraint: Preserve exact overlap semantics and one snapshot for usage metrics
Rejected: Add hourly peaks or truncate intervals | Neither preserves exact concurrency
Confidence: medium
Scope-risk: moderate
Directive: Keep unavailable concurrency distinct from a measured zero
Tested: Rust 1.99.1 rustfmt on changed Rust files; git diff --check; independent source review
Not-tested: PostgreSQL integration, Clippy, and frontend checks pending isolated CI
Co-authored-by: Codex <noreply@openai.com>

* refactor(analytics): limit concurrency changes to unused results

Keep Key usage and Dashboard skip behavior while retaining the existing concurrency calculation and query budget.

Co-authored-by: Codex <noreply@openai.com>

* perf(usage): query insight cards by selected metric

Split health, performance and cost reads by card and selected metric, with independent frontend requests and metric-aware caching. Only the load metric reads concurrency.

Co-authored-by: Codex <noreply@openai.com>

---------

Co-authored-by: Codex <noreply@openai.com>
Co-authored-by: Zyy <zyycn@users.noreply.github.com>
2026-10-30 15:09:24 +08:00
7cd17aff42 feat(usage): add request concurrency trends (#537)
* feat(usage): add request concurrency trends

* perf(usage)!: unify load insights and optimize concurrency

Aggregate concurrency on a single event timeline and reuse lifecycle
indexes instead of joining every request against every time bucket.

Align load summaries and trend points with the four chart metrics,
combine concurrency and scheduling in one view, and right-align the
statistics card selectors. Update the API contract and regression tests.

BREAKING CHANGE: Usage insights concurrency and scheduling P95 fields
move under performance.load and performance.points[].load. Unused
scheduling P50, capacity, and performance coverage fields are removed.

Co-authored-by: Codex <noreply@openai.com>

---------

Co-authored-by: Zyy <zyycn@users.noreply.github.com>
Co-authored-by: Codex <noreply@openai.com>
2026-10-28 17:22:28 +08:00
ZyyandCodex a48139b4c2 fix(store): support literal substring search in admin lists
Match account names and emails, client key names and labels, and observability search fields by literal substrings. Preserve wildcard escaping and existing case rules.

Co-authored-by: Codex <noreply@openai.com>
2026-10-27 16:20:15 +08:00
Zyy c10efea12c perf(dashboard): deduplicate histogram samples and trim unused trend fields
Group identical positive samples before numeric logarithm evaluation while retaining histogram boundaries and weights. Remove nine unused trend fields and synchronize frontend types, preview data, and API documentation.

Validate against the 30-million-row Docker dataset, matching retained statistics and reducing uncached today-summary median latency from 623 ms to 319 ms. Run 133 related tests, Clippy, formatting, frontend checks, and browser tab checks.
2026-10-23 21:06:12 +08:00
ZyyandCodex a23afbf2f5 perf: streamline usage queries and batch account reads
Use literal contains filters for usage names and models, align Key Usage filtering, and remove unused usage query parameters. Optimize summary aggregation, record counting, quota reads, and credential cooldown lookups without adding indexes.

Co-authored-by: Codex <noreply@openai.com>
2026-10-23 01:08:57 +08:00
ZyyandCodex d965bc9050 fix: harden diagnostics and consolidate unreleased migrations
Redact inline credential headers and token assignments in diagnostic exports.
Remove unnecessary usage query copies and the reverse analytics dependency.

Combine unreleased migrations 0032-0035 without changing published migrations.
Update existing migration fixtures and remove five oversized database stress
tests from the regular suite.

Validate with Store, Provider and diagnostics tests, strict Clippy, Rustfmt,
and frontend lint/build.

Co-authored-by: Codex <noreply@openai.com>
2026-10-16 16:55:42 +08:00
ZyyandCodex 8572873c5f refactor(store): standardize database table names
Rename inconsistent tables and their schema objects, update SQL references,
and preserve analytics trigger behavior through the upgrade.

Co-authored-by: Codex <noreply@openai.com>
2026-10-15 23:54:05 +08:00
57c8ef358f feat(settings): configure request history retention (#435)
Allow 1–31 day retention and keep published analytics readable during cleanup.

Co-authored-by: Zyy <zyycn@users.noreply.github.com>
Co-authored-by: Codex <noreply@openai.com>
2026-10-15 22:38:53 +08:00
219888d5c2 perf(store): scale analytics and pagination for large request histories (#423)
Aggregate analytics in PostgreSQL, reconcile synchronous usage with pending source changes, and preserve exact pagination and forecast totals across large histories.

Co-authored-by: Zyy <zyycn@users.noreply.github.com>
Co-authored-by: Codex <noreply@openai.com>
2026-10-13 21:29:25 +08:00
think-backandClaude Opus 5.5 abd60ce60a fix(store): acknowledge analytics versions after projection commits (#404)
The analytics worker acknowledged projected queue versions inside its
repeatable-read projection transaction. Request updates keep rewriting
pending queue rows through the source triggers, and once such a change
commits after the worker snapshot, the acknowledgement row lock fails with
a serialization error instead of skipping the stale version. Under steady
traffic every cycle rolled back the whole projection batch, the queue kept
growing and the worker stayed in backoff.

Commit the projection and publication first, then acknowledge the exact
versions in a separate read-committed transaction. Row locks there
re-check the latest committed version, so rewritten versions no longer
match and stay queued for replay, while locked rows are still skipped.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-12 13:50:02 +08:00
ZyyandCodex cc69368f77 perf(store): scale request analytics and isolate request history (#389)
Build durable request facts and bounded aggregate queries, freeze historical
identities, and move recovery lease updates out of request history rows.
Align token fields across storage, APIs and pages.

Invalidate analytics only after actual request deletions and track embedded
migration inputs during builds.

Co-authored-by: Codex <noreply@openai.com>
2026-10-11 20:57:36 +08:00