Reboot Loop

[Altha Technology]

Hey There,

I’m still experiencing a random reboot loop. I’m having trouble tracking it down. What logs can I check to see who/what is initiate the reboot?

Here is what the docker logs look like:
[services.d] starting services Poste.io administration available at https://10.0.22.101:443 or http://10.0.22.101:80 [services.d] done.2021-12-19 06:35:29 #3760(main) <1c8637>; main; main: rspamd 3.0 is loading configuration, build id: release[!] WARNING: User-initiated shutdown.[cont-finish.d] executing container finish scripts...[cont-finish.d] done.[s6-finish] syncing disks.[s6-finish] sending all processes the TERM signal.

I’m not initiating a shutdown, it just loops forever. I guess, which logs/how can I see what is happening there? Ideas? Could something in the boot process be timing out?

‌

Sincerely,

Jared

[Altha Technology]

Poste Pro was running on version 2.3.4, I rebooted the host, then it starts looping. Have tried rolling back to 2.3.3 and 2.3.1 with no success. Also 2.3.5 no success. Could Redis or something else be storing a cached reboot command or something…?

[Altha Technology]

I don’t know if this is a bug or my issue, would love some guidance on at least how to find what is happening and possibly digging further.

[Altha Technology]

Ok, booting it on a different host allowed it to boot… Maybe something is timing out? The new host is more performant… shrug

[Samuel L. / SSH]

This is a replicable issue.

2.3.4: Mine boot loops a couple of times, then Haraka/nodejs starts giving memory leak warnings relating to Redis or something, More loops until PHP gives max_children errors. More loops, then it “settles“ and the web UI works, but no incoming or outgoing mail.

2.3.5: Doesn’t seem to loop and looks to be a normal boot, but still no functional change.

This is all on Poste Free. Am going to try moving this to a new host next. Edit: New host worked on 2.3.4, but it’s still throwing errors that do not instill confidence.

‌

Edit: After shutting down and starting back up the 2.3.5 container, the boot looping has come back but less and only with one error being logged. Seems to be a Redis issue, possibly related to issue #824. The poste.io site being down is not helping at all.