[TriLUG] need consulting for LAMP server

[Ken MacKenzie via TriLUG]

Another monitoring tool to try out is glances.  It is a python script.  A
pip install of it might be easiest.  Nice output and shows disk IO as part
of its monitoring.

Ken

On Fri, Jun 3, 2016 at 7:09 AM, Ronald Kelley via TriLUG <trilug at trilug.org>
wrote:

> Is your DB setup for replication?  That could cause some issues during
> heavy I/O.
>
>
> Couple suggestions:
>
> During the busy time, run “iotop”, “htop”, and “nmon” in separate windows
> so you can see them together at one time.  These should help pinpoint where
> the contention lies.  Also, suggest you creating a non-DB related HTML page
> to see if you can load just a text page during the busy time.
>
> Try to narrow down the issue to CPU, RAM, Disk, or Network.  While 2x SAS
> drives sounds good, you might be hitting some sort of disk IO wait issue
> causing the DB to slow down.
>
>
>
>
>