Thread: Commitfest Update
So far this commitfest these 10 patches have been marked committed. That leaves us with 175 "Needs Review" and 28 "Ready for Comitter" so quite a ways to go ... * FUNCAPI tuplestore helper function * parse/analyze API refactoring * Add wal_compression=zstd * Add id's to various elements in protocol.sgml * Fix alter data type of clustered/replica identity columns * Fix flaky tests when synchronous_commit = off * Support of time zone patterns - of, tzh and tzm * Optionally automatically disable subscription on error * fix race condition between DROP TABLESPACE and checkpointing * ICU for global collation In case anyone's looking for inspiration... Here are the list of patches marked Ready for Committer: * document the need to analyze partitioned tables * ExecTypeSetColNames is fundamentally broken * pg_dump - read data for some options from external file * Add comment about startup process getting a free procState array slot always * Doc patch for retryable xacts * Function to log backtrace of postgres processes * Allow multiple recursive self-references * Expose get_query_def() * Add checkpoint and redo LSN to LogCheckpointEnd log message * Fast COPY FROM command for the foreign tables * Consider parallel for LATERAL subqueries having LIMIT/OFFSET * Faster pglz compression * Full support for index LP_DEAD hint bits on standby * KnownAssignedXidsGetAndSetXmin performance * Parameter for planner estimates of recursive queries * enhancing plpgsql API for debugging and tracing * Make message at end-of-recovery less scary * Identify missing publications from publisher while create/alter subscription. * Allow providing restore_command as a command line option to pg_rewind * Avoid erroring out when unable to remove or parse logical rewrite files to save checkpoint work * Allow batched insert during cross-partition updates * Add callback table access method to reset filenode when dropping relation * use has_privs_for_role for predefined roles * range_agg with multirange inputs * Add new reloption to views for enabling row level security * Fix pg_rewind race condition just after promotion * Mitigate pg_rewind race condition, if config file is enlarged concurrently. * remove exclusive backup mode -- greg
On Fri, Mar 18, 2022 at 3:08 AM Greg Stark <stark@mit.edu> wrote: > In case anyone's looking for inspiration... Here are the list of > patches marked Ready for Committer: > * Fast COPY FROM command for the foreign tables I have (re-)started reviewing this patch. Best regards, Etsuro Fujita
On Thu, Mar 17, 2022 at 02:07:16PM -0400, Greg Stark wrote: > So far this commitfest these 10 patches have been marked committed. > That leaves us with 175 "Needs Review" and 28 "Ready for Comitter" so > quite a ways to go ... If it were me, I'd move these out of the way somehow; WOA/RWF or move to June: https://commitfest.postgresql.org/37/3291/ Add PGDLLIMPORT to all direct or indirect GUC variables https://commitfest.postgresql.org/37/3568/ PQexecParams binary handling example for REAL data type Is anyone claiming these ? https://commitfest.postgresql.org/37/3142/ Logging plan of the currently running query https://commitfest.postgresql.org/37/3298/ Showing I/O timings spent reading/writing temp buffers in EXPLAIN https://commitfest.postgresql.org/37/3050/ Extended statistics in EXPLAIN https://commitfest.postgresql.org/37/3508/ Avoid smgrimmedsync() during index build and add unbuffered IO API -- Justin
On Tue, Mar 29, 2022 at 4:47 PM Justin Pryzby <pryzby@telsasoft.com> wrote: > If it were me, I'd move these out of the way somehow; WOA/RWF or move to June: > > https://commitfest.postgresql.org/37/3291/ Add PGDLLIMPORT to all direct or indirect GUC variables I plan to do this yet, but it seemed best to leave it until the end to avoid creating as many merge conflicts. -- Robert Haas EDB: http://www.enterprisedb.com