January 8, 2016 at 4:42 pm #13475
After running in a fairly stable way for months, we’ve upgraded to MONyog 6.5 before Christmas, and had the process die 3 times in 4 weeks. Each time the process logged a segmentation fault. Running on Centos 6.5:
[root@(server) MONyog]# cat /etc/centos-release
CentOS release 6.5 (Final)
[root@(server) MONyog]# uname -a
Linux (server.fqdn) 2.6.32-431.3.1.el6.x86_64 #1 SMP Fri Jan 3 21:39:27 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux
[root@(server) MONyog]# rpm -qa | grep MONyog
The error was logged 3 minutes after a failed call to the AWS API to get logs from an RDS server, and a curl timeout seems to be involved in the crime. It had been logging a similar failure every 5 minutes; something about that RDS DB isn’t happy – but that shouldn’t be crashing my entire database monitoring application!
[6.5] [2016-01-07 18:55:59] [Server: (endpointname)] populatemysql.cpp(471) ErrCode:-1 ErrMsg:RDS Error log not present
[6.5] [2016-01-07 18:58:56] linservicemgr.cpp(106) ErrCode:11 ErrMsg:Stopping MONyog: Received signal SIGSEGV -- Segmentation fault!
We’re running the enterprise version, if that’s relevant. I’ll try to hunt down the customer login details and send over the core dump, but wanted to post something public so anyone else with the same issue can see it’s not a one-off!
January 19, 2016 at 12:07 pm #35619
This issue related to the curl library crash while resolving too many hostnames was fixed in our latest release MONyog-6.51. You can go ahead and upgrade and the issue will be fixed.
January 19, 2016 at 12:19 pm #35620
You must be logged in to reply to this topic.