Tim Düsterhus

> Not sure if it's entirely useful, but here it is It at the very least indicates which thread triggered the watchdog. I've updated the issue title and will defer...

> I'll try to find a way to the owner of https://launchpad.net/~vbernat/+ppa-packages to see if there's any way to improve this situation for the arm packages This is Vincent Bernat...

> FATAL: bug condition offset >= qcs->tx.offset" matched at src/mux_quic.c:937 Just to clarify: This is not a regular segfault, but rather an assertion failure that intentionally crashes the process by...

> Just to complement, it's not a pointer dereferencing but a BUG_ON statement on an equality comparison. Ah, sorry if I said something confusing there :smiley: I meant to say...

For reference: https://www.mail-archive.com/[email protected]/msg42745.html

@wtarreau Should this still happen?

@wtarreau Conference is coming up, I know: I just remembered this issue and wanted to bump it before I forgot about this myself again :smiley:

# Clean start ``` $ systemctl restart haproxy $ echo show proc |nc -U /run/haproxy-master.sock # 1275 master 0 [failed: 0] 0d00h00m01s 2.6.2-1~bpo10+1 # workers 1277 worker 0 0d00h00m01s 2.6.2-1~bpo10+1...

# data from the master socket ``` $ echo show proc |nc -U /run/haproxy-master.sock # 1275 master 1 [failed: 0] 0d00h01m48s 2.6.2-1~bpo10+1 # workers 1708 worker 0 0d00h01m22s 2.6.2-1~bpo10+1 #...

Okay, I've evacuated the node in question with a `monitor fail if TRUE`. It does not serve any user traffic. # Immediately after systemctl restart haproxy ``` $ echo show...