On 12/13/2014 12:55 PM, Edson Carlos Ericksson Richter wrote:
> Dear list,
>
> I've been searching in web for guidelines on OS (Linux) and PostgreSQL (9.3.5) active monitoring best practices.
> Can someone share experiences?
> I'm inclined to look at Cacti and Nagios. Any other experiences? Recommended books?
> I don't want to use SaaS for monitoring - I'll have a cloud server hired specifically for this purpose, outside my
maindata center infrastructure.
>
> Thanks in advance,
>
> Edson
>
>
Stats are one thing, but errors are another. I've found my best monitor is rsyslog and a perl script.
rsyslog.conf contains:
local0.* action(type="omprog"
binary="/usr/local/bin/logMonitor.pl"
template="RSYSLOG_TraditionalFileFormat")
the perl script is sort of like:
while (<>)
{
emailme() if (/error/);
}
-Andy