Pre-production hardening: schema, queries, ingestion, and API#665

Closed

aditya1702 wants to merge 17 commits into

blend/pr6-integration-testsfrom

pre-prod-hardening

Commits on Jul 16, 2026

fix(graphql): group dataloader batches by query shape
aditya1702
committed
fix(ingest): graceful shutdown on SIGTERM with clean resource teardown
aditya1702
committed
test(graphql): lock freighter query complexity under the prod limit
aditya1702
committed
feat(db): retune storage params, API-sort indexes, idempotent TSDB policy config
aditya1702
committed
perf(data): time-pinned lateral probes for batch queries; query hygiene
aditya1702
committed
feat(data): NUMERIC amount columns; deterministic namespaced state_change IDs
aditya1702
committed
refactor(ingest): move classification and metadata RPC out of DB transactions
aditya1702
committed
fix(ingest): split-brain defenses, permanent-error fast fail, bounded pools
aditya1702
committed
feat(graphql): API hardening — timeouts, depth limit, error masking, root time bounds
aditya1702
committed
docs(ops): production operations reference; TimescaleDB 2.28.2 pins; README sync
aditya1702
committed
feat(db): chunk skipping on transactions.ledger_number for windowed gap detection
aditya1702
committed
remove the old loadtest backend
aditya1702
committed
fix(ingest): exempt the checkpoint load transaction from idle-in-transaction timeout
aditya1702
committed
perf(db): consolidate hypertable indexes into query-ordered primary keys
aditya1702
committed
perf(graphql): 1ms dataloader collection window; per-loader batching metrics
aditya1702
committed
docs(ops): drop the TimescaleDB upgrade runbook
aditya1702
committed
perf(db): NUMERIC columns and tuned storage for blend_claimed tables
aditya1702
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Pre-production hardening: schema, queries, ingestion, and API#665

Pre-production hardening: schema, queries, ingestion, and API#665
aditya1702 wants to merge 17 commits into
blend/pr6-integration-testsfrom
pre-prod-hardening

Commits on Jul 16, 2026

fix(graphql): group dataloader batches by query shape

fix(ingest): graceful shutdown on SIGTERM with clean resource teardown

test(graphql): lock freighter query complexity under the prod limit

feat(db): retune storage params, API-sort indexes, idempotent TSDB policy config

perf(data): time-pinned lateral probes for batch queries; query hygiene

feat(data): NUMERIC amount columns; deterministic namespaced state_change IDs

refactor(ingest): move classification and metadata RPC out of DB transactions

fix(ingest): split-brain defenses, permanent-error fast fail, bounded pools

feat(graphql): API hardening — timeouts, depth limit, error masking, root time bounds

docs(ops): production operations reference; TimescaleDB 2.28.2 pins; README sync

feat(db): chunk skipping on transactions.ledger_number for windowed gap detection

remove the old loadtest backend

fix(ingest): exempt the checkpoint load transaction from idle-in-transaction timeout

perf(db): consolidate hypertable indexes into query-ordered primary keys

perf(graphql): 1ms dataloader collection window; per-loader batching metrics

docs(ops): drop the TimescaleDB upgrade runbook

perf(db): NUMERIC columns and tuned storage for blend_claimed tables