PostgreSQL

14 March 2025 22:14

I have patroni setup with 3 nodes, master sync async, patroni conf slot is true and pg has 2 phisycal rep slots, replicas seems working well, problem is vacuum not removing dead tuples because rep slots keeps xmin far behind from current, what can I do, how to solve this xmin problem

PostgreSQL

12 March 2025 17:34

It's very good you have other consistency support methods beyound database. I'm always having to do some levels of recovery when I'm going to implement some busyness-critical apps.
I just like the relational database to be one of those levels, not a simple throw-out unreliable storage!

PostgreSQL

12 March 2025 17:19

Most dreaded data loss scenario is actually an edge case where you cancel update/insert/delete statement waiting on syncrep before failover. We made application queries re-run friendly and after autofailover we would just stream last 5 minutes of events all over again (since it was an oltp with no queries greater than 5min allowed.) making data loss impossible.

https://github.com/patroni/patroni/pull/1414/files

PostgreSQL

12 March 2025 16:04

Damn my workplace gives 48 core xenon server

PostgreSQL

12 March 2025 15:24

Hi pls suggest load balancer with failover

PostgreSQL

12 March 2025 14:38

There are no good autofailovers for postgres.

PostgreSQL

12 March 2025 09:51

pg_dump reads your database logically and converts the data, so it will be "slow"

PostgreSQL

12 March 2025 09:31

https://b-peng.blogspot.com/2022/03/pgpool-debian.html?m=1

PostgreSQL

12 March 2025 09:19

Use pg_backup_start to freeze your data folder and then use scp or rsync or cp or whatever to clone your data folder as fast as possible

PostgreSQL

12 March 2025 09:11

does anyone have any advice on exporting large (not huge) PostgreSQL databases in Python with Subprocess? I am doing this sort of thing:

dump_process = subprocess.Popen(
    dump_command,  # /usr/bin/pg_dump etc...
    stdout=subprocess.PIPE,
    stderr=subprocess.PIPE,
    env={"PGPASSWORD": "mypw"}
)

dump_result = dump_process.wait(timeout=300)

PostgreSQL

12 March 2025 07:09

Having you any document for implementation and any blog related

PostgreSQL

12 March 2025 05:46

My company also suggest me but I am not sure to auto failover in pg pool support or not?

PostgreSQL

12 March 2025 05:38

Repmgr can manage load balancer?

PostgreSQL

11 March 2025 17:28

Anyone going to PG Orlando?
Anyone have a Promo Code?

PostgreSQL

11 March 2025 13:00

Hello, this kind of support warrants Amazon's action. We can support you only on open source Postgresql®

PostgreSQL

13 March 2025 03:42

Finally got around implementing this awesome query (also thanks for making me understand lateral joins) (I know its been a long time 🙃)

Thanks once again to

- @Intrbiz for mentioning lateral joins
- @unfoxo for not exists (which is so good i can't stop using it)
- @dear_tomato for the query

PostgreSQL

12 March 2025 17:28

Objection. I'm just fine with the other transaction modes (especially when the programmer understands consequences)!

PostgreSQL

12 March 2025 17:02

For the record: Ilya has pretty distinct opinions on basically everything that is not a bare-metal server running no HA and does every transaction in SERIALIZABLE mode.

Others, including myself, have made good experiences with Patroni. Those included putting Patroni on top of an already running replicated PG cluster with no downtime.

PostgreSQL

12 March 2025 15:40

If the user if fine to throw out some transactions or lay down several hours while DBAs are getting along the patroni... Then yes, it might be kind of a good... But why do they want autofailover then?

PostgreSQL

12 March 2025 15:04

Depends on the definition of good. Good from DBA perspective or End user perspective. 3-node patroni HA cluster is great from user perspective but not so good from dba ease of setup/maintenance perspective.

PostgreSQL

12 March 2025 09:52

copying from the hdd goes as fast as the storage obviously

PostgreSQL

12 March 2025 09:49

are you saying the data directory is the bottleneck?

PostgreSQL

12 March 2025 09:19

You can use pg_lscluster -j to get json info on every database you're running

PostgreSQL

12 March 2025 09:18

Just as i was writing the same exact thing yesterday, haha

PostgreSQL

12 March 2025 07:10

Pls guide and suggest

PostgreSQL

12 March 2025 05:51

It supports auto failover

PostgreSQL

12 March 2025 05:40

Pgpool-II also meet ur requirement

PostgreSQL

12 March 2025 05:00

Hi everyone I want setup load balancer with auto failover can you suggest which tool is good if any document pls guide me in postgresql version 14

PostgreSQL

11 March 2025 13:25

Most likely you need support from AWS. But "I need support" is a very broad request, you need to be more specific.

PostgreSQL

11 March 2025 04:44

Hello All

(currently we are doing migration db2 to PostgreSQL using AWS SCT)

I need support, Could you please help me on this