Haolin Wang

Vacuum index on Append-Optimized table enhancement backport.

1

This is back-porting VACUUM indexes enhancement https://github.com/greenplum-db/gpdb/pull/13255 In the past, we rely on the visibility map to check if an index entry can be removed. Scanning VM is not such...

Enable index only scan on ao/aocs table.

2

The index scan has disabled on ao/aocs table for that random access performance on the ao/aocs table is extremely low. But index only scan does not access data files. Its...

This is backport of https://github.com/greenplum-db/gpdb/pull/13891. Commits start blocking only if STREAMING or CATCHUP within range. In real scenarios it often happens that the mirror node is down (probably not really...

Enable catchup within range on segment.

1

Commits start blocking only if STREAMING or CATCHUP within range. In real scenarios it often happens that the mirror node is down (probably not really down) and out-of-sync for some...

Add throttle logic to slow down WAL writes in CATCHUP phase.

4

There was a guc repl_catchup_within_range that "Sets the maximum number of xlog segments allowed to lag when the backends can start blocking despite the WAL sender being in catchup phase."...

Truncate operation interleaves between END_OF_RECOVERY and CHECKPOINT may panic for AO table.

7

## Bug Report ### Greenplum version or build 6X ### OS version and uname -a Linux ### autoconf options used ( config.status --config ) ### Installation information ( pg_config )...

[7X] Implement aovisimap cache

7

This is a followup work for PR https://github.com/greenplum-db/gpdb/pull/16603 to implement an aovisimap cache for optimization. The idea is reusing `VfdCache` caching mechanism, using segno+firstrowno as a cache key to hash...

[6X] Fix TRUNCATE could remove shared AO hash entry in use by other transa…

6

…ctions. An example repro is: -- session1, under testdb1: ``` testdb1=# create table aorow(a int) with (appendonly=true); testdb1=# insert into aorow select generate_series(1,10); testdb1=# create database testdb2 with template testdb1;...

Sampling on CO tables requires all columns sub-segfiles have the same total `tupcount` presented physically, ADD COLUMN operation didn't populate physical rows which have default values for better performance. This...