Backend robustness for 100-200 services - #24

Merged

aksOps merged 20 commits into

mainfrom

feat/backend-robustness-100-200-services

Apr 23, 2026

Commits on Apr 23, 2026

fix(graphrag): propagate real span Status into OnSpanIngested

aksOps
and
claude
committed
1b222e0
feat(graphrag): expose event-drop metric + configurable workers/queue

aksOps
and
claude
committed
75f31c6
fix(api): exempt OTLP /v1/* from per-IP rate limiter

aksOps
and
claude
committed
0d0f583
docs(api): document /v1/* rate-limit exemption trade-off

aksOps
and
claude
committed
bbb0833
feat(grpc): apply message size, stream, and keepalive limits

aksOps
and
claude
committed
fb85130
feat(config): bound GRPC_MAX_RECV_MB and GRPC_MAX_CONCURRENT_STREAMS

aksOps
and
claude
committed
3e705f3
feat(retention): parallelize purges + tune batch size/pacing

aksOps
and
claude
committed
51852c0
fix(retention): recover panics in parallel purge goroutines

aksOps
and
claude
committed
7d1b99f
feat(storage): refuse SQLite in APP_ENV=production without override

aksOps
and
claude
committed
7703438
feat(telemetry): expose DB pool stats

aksOps
and
claude
committed
7709782
chore(deps): tidy go.mod after Task 7 telemetry test imports

aksOps
and
claude
committed
93b9778
feat(dlq): export eviction metrics

aksOps
and
claude
committed
8111d60
feat(graphrag): dedup investigations with 5m cooldown

aksOps
and
claude
committed
a6ef0d5
perf(storage): driver-switched p99 computation

aksOps
and
claude
committed
65f3742
fix(storage): address Task 10 dashboard p99 review findings

aksOps
and
claude
committed
0e4da47
feat(loadsim): add 200-service OTLP load simulator

aksOps
and
claude
committed
6b63e0b
fix(loadsim): address review findings

aksOps
and
claude
committed
8cccd14
test(storage): shrink p99 cap in test to survive -race

aksOps
and
claude
committed
e137451
docs(operations): document Tasks 1-11 metrics and load simulator

aksOps
and
claude
committed
3e64f1a
docs: address final review findings

aksOps
and
claude
committed
8e85a85

Navigation Menu

Uh oh!

Backend robustness for 100-200 services - #24

Backend robustness for 100-200 services#24
aksOps merged 20 commits into
mainfrom
feat/backend-robustness-100-200-services

Commits on Apr 23, 2026

fix(graphrag): propagate real span Status into OnSpanIngested

feat(graphrag): expose event-drop metric + configurable workers/queue

fix(api): exempt OTLP /v1/* from per-IP rate limiter

docs(api): document /v1/* rate-limit exemption trade-off

feat(grpc): apply message size, stream, and keepalive limits

feat(config): bound GRPC_MAX_RECV_MB and GRPC_MAX_CONCURRENT_STREAMS

feat(retention): parallelize purges + tune batch size/pacing

fix(retention): recover panics in parallel purge goroutines

feat(storage): refuse SQLite in APP_ENV=production without override

feat(telemetry): expose DB pool stats

chore(deps): tidy go.mod after Task 7 telemetry test imports

feat(dlq): export eviction metrics

feat(graphrag): dedup investigations with 5m cooldown

perf(storage): driver-switched p99 computation

fix(storage): address Task 10 dashboard p99 review findings

feat(loadsim): add 200-service OTLP load simulator

fix(loadsim): address review findings

test(storage): shrink p99 cap in test to survive -race

docs(operations): document Tasks 1-11 metrics and load simulator

docs: address final review findings