[SOLVED] decryption failed or bad record on remote sync

[DrillSgtErnst]

Hi,
I am syncing three servers over the internet.
on one I have issues.
The sync is having problems as follows

2024-04-12T21:17:35+02:00: sync group vm/100 failed - error:0A000119:SSL routines:C:decryption failed or bad record mac:../ssl/record/ssl3_record.c:622:

The error occurs sometimes after one hour, sometimes after 3-4

upon restart I get
2024-04-15T08:53:00+02:00: re-sync snapshot vm/100/2024-04-10T20:00:01Z
2024-04-15T08:53:00+02:00: no data changes
2024-04-15T08:53:00+02:00: percentage done: 33.33% (1/3 snapshots)
2024-04-15T08:53:00+02:00: sync snapshot vm/100/2024-04-11T20:00:02Z

Is this eventually just an disconnection issue?

 

[ggoller]

Hi!
this is probably related to a flaky network. Check out these forum threads/issues:
- https://github.com/openssl/openssl/issues/11727
- https://forum.proxmox.com/threads/pbs-sync-job-fails-with-ssl-error-after-8min-up-to-30mins.113311/
- https://forum.proxmox.com/threads/pbs-sync-failed-each-time.113921/#post-492989

 

[DrillSgtErnst]

Well the error Message "e1000e reset adapter unexpectedly" indeed appeared on my server too. I wasn't expecting too much of this, but this seemingly was the problem. I applied the fix via ethtool like your colleague Thomas reported

ethtool -K ens18 tso off gso off gro off


To apply it automatically on boot edit /etc/network/interfaces and add a post-up entry for that interface (can also add it on the bridge):
iface ens18 inet manual
# ... existing lines
post-up /sbin/ethtool -K ens18 tso off gso off gro off

Seems to work for now


EDIT:
I can confirm it was the problem. Worked for tha last days. I marked this as solved.