[LON-CAPA-admin] LON-CAPA Performance Monitoring
Bynum, Lee Hamilton
leebynum at illinois.edu
Tue Oct 4 14:04:04 EDT 2016
At Illinois we've used CPU load and memory usage as indicators of performance and have found that between them we have a sense of when our resources are being stressed. User load has also been a good indicator, but when things get really busy it can be misleading. If you've heard of any other good metrics by which to measure poor performance in LON-CAPA, though, I would be very interested. Defining what is and is not acceptable levels of performance during peak usage has been an ongoing target for us.
> -----Original Message-----
> From: lon-capa-admin-bounces at mail.lon-capa.org [mailto:lon-capa-admin-
> bounces at mail.lon-capa.org] On Behalf Of Lucas, Mark
> Sent: Monday, September 12, 2016 9:31 AM
> To: LON-CAPA Admin
> Subject: [LON-CAPA-admin] LON-CAPA Performance Monitoring
> I need some advice.
> Over the next half year I will be transitioning our LON-CAPA servers over to
> Ohio University IT group. They will be placing the servers on new hardware,
> presumably virtual machines.
> I’ve been fortunate until now. I keep multiple generations of machine
> which means I have a decent number of access servers to take up the slack
> at peak times. They will presumably be looking to try to minimize the
> they need to apply to LON-CAPA, so I foresee an issue being some form of
> metric at peak times that can be used to negotiate hardware needs.
> Does anyone have advice on such a metric and what might be considered
> a threshold that suggests poor performance?
> I’m presuming something like page response time would be best.
> Does anyone do any monitoring of such a parameter?
> Mark Lucas email:
> lucasm at ohiou.edu
> 252D Clippinger Lab phone:
> Department of Physics and Astronomy fax: (740)593-0433
> Ohio University
> Athens, OH 45701
> LON-CAPA-admin mailing list
> LON-CAPA-admin at mail.lon-capa.org
More information about the LON-CAPA-admin