I work for a consulting / hosting firm in Topeka, and I have a server
that is periodically crashing. The last couple of times, it appears to
have been due to the server running out of memory. It is a virtual
server (on VMware ESX), so I increased the RAM from 2GB to 4GB and the
virtual CPUs from 1x single-core to 2x quad-cores. However, it may be a
week or more before I can tell if this has resolved the issue.
What I am looking for is advice on how to track down the cause
definitively, and prevent it from occurring again. (My client is
complaining about the downtime.) When this occurred last time (~ a week
ago), I installed sysstat to try and track the performance, but when I
went to look at today's report, I got a message saying: "Invalid system
activity file". Viewing yesterday's report works just fine, so I'm
thinking that the hard-reset I performed this morning F'd up the report.
We would be willing to pay the right person for their expertise, and I
hope to learn something in the process. I'm fairly knowledgeable, but I
think I may be in over my head on this one.
The OS is CentOS 6.3, and it is dedicated to a single website running
the Magento e-commerce software, which relies heavily on MySQL. Any
advice or assistance would be appreciated.
Thanks in advance,
~ j.