PostgreSQL

14 February 2026 08:50

Is there any serious use case for hash partitioning apart from sharding? 🤔

PostgreSQL

14 February 2026 00:07

is this possible to convert a big table into hash partitioned on-line?

PostgreSQL

13 February 2026 06:54

Hi anybody knows about pgbouncer

PostgreSQL

10 February 2026 15:47

This place is for postgresql, not Oracle.

PostgreSQL

10 February 2026 14:42

It's not received the hype coverage it deserves... also, the list of filesystems actually supporting it is rather short.
And the rsync trick for replicas doesn't work with it AFAICS, so it's not for everyone.

PostgreSQL

10 February 2026 12:32

@unfoxo thanks!

PostgreSQL

10 February 2026 12:19

link means it will create the new upgraded version with hardlinks to the old datafiles? so it will modify the older version?
copy is the safest but it takes longer to copy the hole data?

PostgreSQL

10 February 2026 11:59

I use --link because my filesystem supports snapshots, so i can go back to the old snapshots and undo the upgrade if anything happens. But if something bad happens, the files that were changed by the newer postgresql version during the upgrade will not work on the older one.

PostgreSQL

10 February 2026 11:56

Do you use the

--link

pg_upgrade

PostgreSQL

05 February 2026 10:35

Ohhh ok thank you but alternatively postgresql service also stopped at that same time.

PostgreSQL

05 February 2026 10:30

At the same time all services one time stopped at a specific time like 6 am to 6:15am

PostgreSQL

05 February 2026 10:27

2026-02-05 06:12:37.305 IST [1059482<><><><>]>:LOG: background worker "logical replication launcher" (PID 1059504) exited with exit code 1
<2026-02-05 06:12:37.428 IST [1059498<><><><>]>:LOG: shutting down
<2026-02-05 06:12:37.439 IST [1059498<><><><>]>:LOG: checkpoint starting: shutdown immediate
<2026-02-05 06:12:37.457 IST [1059498<><><><>]>:LOG: checkpoint complete: wrote 1 buffers (0.0%); 0 WAL file(s) added, 0 removed, 0 recycled; write=0.012 s, sync=0.002 s, total=0.020 s; sync files=1, longest=0.002 s, average=0.002 s; distance=6 kB, estimate=1841 kB; lsn=72/4BB2DE48, redo lsn=72/4BB2DE48
<2026-02-05 06:12:37.680 IST [3439086<10.31.xxx.42><walreceiver><replication><[unknown]>]>:streaming 72/4BB2DEC0LOG: disconnection: session time: 0:08:23.275 user=replication database= host=10.31.xxx.42 port=58424
<2026-02-05 06:12:38.674 IST [1059482<><><><>]>:LOG: database system is shut down

PostgreSQL

05 February 2026 10:07

You'll need to check/share the logs!

PostgreSQL

04 February 2026 13:24

Your Monitoring has an opinion on that... CPU, Memory, Disk saturation, network saturation.

PostgreSQL

03 February 2026 02:06

thats "better". at least it publicly open. 🙈

we cant imagine how massive leaked credential within claude, angravity, cgpt. it kept privately.

is it really disposed?. we dont know. is it synced to authority espionage db ?. we dont know either.

PostgreSQL

14 February 2026 01:49

Strictly speaking: yes. But it would require you to build the hashed table in the background. When you are ready you lock both tables and swap the names.

I would strongly advise to rethink your partitioning set up. When you use hash partitioning you can get a lot of full page writes on your index because of the random inset positions. It is also very hard to expand the number of partitions in the future.

I would recommend range based partitioning. Maybe investigate the uuidv7 data type for your partitioning column.

PostgreSQL

13 February 2026 06:59

I know a little bit, yes

PostgreSQL

13 February 2026 02:56

Looking for Postgre sql DBA support on Monday 2pm -2:45pm CST

PostgreSQL

10 February 2026 15:47

Looking for Oracle DBA opportunity with 9.5+ years of experience.... Please let me know if any openings

PostgreSQL

10 February 2026 14:20

There's also --clone now, for some filesystems. Combines the best of both worlds.

PostgreSQL

10 February 2026 12:23

Since the postgresql on-disk format is mostly static, most files will just be hardlinked

PostgreSQL

10 February 2026 12:00

The safest is using copy, it'll also as a bonus check if all the files are okay and refresh bitrot on ZFS pools / RAID arrays 😺

PostgreSQL

10 February 2026 11:58

Depends. If you use --link, you cannot go back to the old version if something wrong happens.

PostgreSQL

05 February 2026 11:28

What you have been told is likely the case, it wouldn't just shut down for no reason. In addition to the above, do you have an unattended-upgrades package installed? apt and other package managers may shut down and restart services if required to by the install procedure. Check also if your host has a particular maintenance policy or notices.

PostgreSQL

05 February 2026 10:34

The logs show a proper shutdown indeed. So it's SystemD.
Probably some smartass ansible or puppet thing believing "this requires a restart of all services".
Talk to your OS admins (if that's not yourself, obviously)...

PostgreSQL

05 February 2026 10:28

is there anything else?

PostgreSQL

05 February 2026 10:07

I highly doubt that this is not triggered by something on OS level. You'd be the first person globally to experience random restarts of PG... except for Signal 11 or similar problems.

PostgreSQL

05 February 2026 10:01

I am getting issue on postgres server automatically restarted for all services and postgresql also restarted on specific dates like 30 jan and feb-5. How to fix it pls help me

PostgreSQL

04 February 2026 12:34

Our ERP has 4 modules which r in 4 tomcat...is it ok to migrate to one tomcat??

PostgreSQL

03 February 2026 01:39

https://www.linkedin.com/pulse/ai-agents-concerning-agency-moltbook-savneet-singh-iazje/

Lol well deserved Ai-slop hack