Franck Pachot

[Docs] docker-compose.yaml example not working in some enviromnents

3

### Description Two issues were reported regarding the example docker-compose.yaml in https://docs.yugabyte.com/latest/deploy/docker/docker-compose/ - port bind error (see https://forum.yugabyte.com/t/docker-compose-connection-refused/1319/3). Solution: add hostname - `FATAL: lock file "/tmp/.yb. /.s.PGSQL.5433.lock" already exists` after...

[YSQL] Slow update on INCLUDE index (1 flush per write)

3

Jira Link: [DB-9891](https://yugabyte.atlassian.net/browse/DB-9891) ### Description There are two ways to create a covering index: add columns in INCLUDE or add them at the end of the key. INCLUDE is supposed...

Hi, There's an anomaly in PostgreSQL that is not caught by those tests because it doesn't implement statement restart like Oracle or YugabyteDB (I didn't test other MVCC). Here is...

Bug: Can't detect useHostColumn when connecting with hostname rather than IP

1

- [X] bug report - [ ] feature request When we connect with hostname rather than IP addess we get: `WARNING: TopologyAwareLoadBalancer: Either private or public should have matched with...

New region opened this week

Jira Link: [DB-11374](https://yugabyte.atlassian.net/browse/DB-11374) ### Description The following starts a transaction and a row is inserted by another transaction. This rows should not be visible by the transaction started before it:...

Jira Link: [DB-10295](https://yugabyte.atlassian.net/browse/DB-10295) ### Description A user reported [[YSQL] Transactional COPY FROM consumes excessive amount of memory](https://github.com/yugabyte/yugabyte-db/issues/21370) Here is an easy way to reproduce. ``` create table demo_ref ( id...

"ERROR: SET TRANSACTION ISOLATION LEVEL must not be called in a subtransaction" with autosave

2

**Describe the issue** With autosave=always, a SET TRANSACTION ISOLATION LEVEL at the beginning of the transaction fails with `ERROR: SET TRANSACTION ISOLATION LEVEL must not be called in a subtransaction,...

DISCARD SEQUENCES is useless and not scalable

3

On connection reset, DISCARD SEQUENCE is [called](https://github.com/npgsql/npgsql/blob/87aeafbc8c68f9f31d3b80e68b072475ba910768/src/Npgsql/Internal/NpgsqlConnector.cs#L2201). Why discard the sequence cache? This action eliminates all the benefits of having a sequence cache (the same as CACHE 1) while introducing...

Wrong logic for InsertOrUpdate

4

It seems that InsertOrUpdate generates a command with two statements: ``` [info] 2024-06-15T18:34:30.527 [-2] DEBUG slick.jdbc.JdbcBackend.statement Preparing statement: update "test" set "value"=? where "id"=?; insert into "test" ("id","value") select ?,?...