All Murmur Servers Crashed Simultaneously

[mumblenz]

I have been running mumble servers for years, and have always had many servers running at once on the same box with no issues. Firstly I'll give you a brief overview of my setup, then ill go into the problem a bit. Any insights at all would be greatly appreciated.

 

• Each server has its own murmur.x86 file and config file.

• Each server is located in its own User folder & runs as said user

• Each server has its own database and log files, also in their own User folder

• Each server runs on a different port and or IP

• Some servers are public and some are private

• The server was under almost no load at the time of the crash

• Also the network logs show extremely minimal traffic leading up to the crash

• The server uptime is 60 ish days and no other processes crashed as far as I can tell

• All servers are running version 1.2.4 on Debian 6.0.7

 

Everything had been running for 60 ish days without issue (and before that about 8 months). Then every single murmur server crashed at exactly the same time. The murmur logs just stop, the is no errors in the logs for the few hours leading up to the crash. Also there were no common users connecting to multiple servers (rule out a user taking them down in some way), also some servers are private password protected servers requiring authd user certs to access them at all. The syslogs show nothing odd, no root logins during that time frame, no other services stopping randomly.

This has me totally baffled. As I said before any insights at all would be greatly appreciated.

[kissaki]

I can not think of anything Murmur controlled here.

Do you bind Murmur to DBus?

A pkill murmur.x86 would have been logged?

Maybe the parent process crashed? Is it a common parent?