Precise wait-event timing and tracing
needs reviewLow-overhead core timing, trace capture, coverage analysis, and independent observer-effect benchmarks.
postgres@hacking:~$ cat WORK
Patch lines.
What is moving, what needs review, what is blocked, and where the durable artifacts live.
4already committed upstreamAlready in Postgres and available to everyone
[merged]transaction_timeoutStop transactions that remain open longer than the configured limit.[merged]bounded psql \watchLet psql users stop \watch automatically after a fixed number of executions.[merged]UUIDv7 generationTurn the original ULID exercise into PostgreSQL's sortable uuidv7() implementation.[merged]MultiXact standby deadlock fixFix MultiXact creation, deadlock, and corruption edge cases on standbys.
open
Log the correct commit-record LSN for DROP TABLE and related destructive operations to make PITR recovery less guessy.
Keep logical slots usable on a standby fed by WAL archives by pausing and safely resuming recovery around conflicts.
Partial materialized-view refresh
needs reviewAdd WHERE support to REFRESH MATERIALIZED VIEW with explicit concurrency semantics.
Whole-WAL-record compression
blockedCompress or batch larger WAL records, originating in B-tree build experiments.
Custom protocol command tags
explorationNegotiate custom command-completion tags through a _pq_ protocol capability.
Merge sparse sibling pages to reduce index bloat without rebuilding the entire index.
also supporting
TRACE / wait-event tracing
Dmitry FominDmitry’s pg_wait_tracer and core wait-event timing work, supported here with design sessions, reviews, coverage analysis, and independent benchmarks.
pg_stat_statements scalability
GSoC 2026 + AWSKirk and Nik mentor Quan Hoang Truong’s GSoC project. A related AWS-led patch series now implements a broader pgstat-based redesign for the same contention problem.
known blockers
Precise wait-event timing and tracingNeeds more reviewer cycles and benchmark scrutiny in CommitFest.Recovery landmarks for destructive DDLThe corrected PR changes have not yet been posted back to -hackers and CommitFest.Logical decoding from archive-fed standbysCrash-safety and promotion behavior need complete end-to-end coverage before upstream submission.Partial materialized-view refreshCommitFest entry currently needs review.Whole-WAL-record compressionClosed CommitFest entry is waiting on the author.Custom protocol command tagsNo CommitFest entry and the protocol design needs broader consensus.B-tree bloat removal (page merge)Prototype correctness and concurrency design are not ready for CommitFest.
other public outcomes
unqualified DML guardspatch proposed · Prototype server-side protection against UPDATE or DELETE without qualification.pg_stat_kcache observer-effect fixpatch proposed · Reduce self-hidden I/O accounting caused by extension locking.whole-WAL-record compressionCommitFest patch · Compress or batch large WAL records, originating in B-tree build work.recovery landmarks for destructive DDLCommitFest patch · Record a useful commit LSN for recovering from DROP TABLE and related operations.ANALYZE steps in dump outputpatch proposed · Preserve or trigger statistics collection during dump and restore workflows.B-tree bloat removal (page merge)patch proposed · Merge sparse sibling pages to reduce index bloat without a full rebuild.partial materialized-view refreshCommitFest patch · Add WHERE support to REFRESH MATERIALIZED VIEW.precise wait-event timing and tracingCommitFest patch · Publish low-overhead timing, trace capture, coverage analysis, and benchmarks.custom protocol command tagspatch proposed · Negotiate command-completion formats through the _pq_ protocol capability.pglz compression experimentpublic experiment · Build and publish an agent-assisted testbed for improving pglz compression.logical decoding from archive-fed standbyspatch series · Pause and safely resume recovery instead of invalidating conflicting logical slots.
shipped work lines
MultiXact standby deadlock
shippedFix standby MultiXact creation and deadlock/corruption edge cases.