@Max Kostikov which query is it that took so long?

2026-10-21 04:22:58

adminsforum@hubzilla.org

Markus wrote the following post Wed, 21 Oct 2026 04:22:48 -0700

Hubzilla isn't running well right now. The database and, at times, the queue are running significantly slower than usual. Have you noticed this, too?

2026-10-21 04:44:08

Saiwal

sk@utsukta.org

on my hub enabling pubstream adds a lot to cpu load which contributes to this, irrespective of what theme i use. So disabling pubstream or keeping it local only helps.

2026-10-21 05:37:15

Alfred Bühler

abu@hub.alfredbuehler.ch

@Hubzilla Support Forum The speed is as usual here. But last night my DB size increased by 54 MB, which is significantly higher as usual.

2026-10-21 06:51:29

*jayrope

jrp@hub.kliklak.net

@Hubzilla Support Forum I am on 11.4.2 and all systems are running well. No database or pubstream related troubles either. The only queues I see regularly piling up and then go down again are mostly from other Hubzilla servers. The rest of the Fediverse seems to not do this.

@Markus @Alfred Bühler @Saiwal

2026-10-22 02:29:58

Max Kostikov

kostikov@zotum.net

@Hubzilla Support Forum
If you are under reliatively fresh MySQL 8 then try to add this index
ALTER TABLE workerq ADD INDEX idx_claim (workerq_reservationid, workerq_priority DESC, workerq_id);
In my case, on the `dev` branch of Hubzilla, this significantly improved the performance of the queueworker from tens of seconds to a couple hundred milliseconds.
UPD: CPU usage has also dropped sharply.

FYI @Mario Vavti