On 05/14/2012 11:54 AM, Paul Libbrecht wrote:
Is there really no-one using any monitoring tools? How is the myxwiki.org farm monitored?
Yep, we're doing monitoring, but the sysadmins are the only ones that can answer that question. They probably haven't notice this thread yet. Did you see https://github.com/xwiki-contrib/xwiki-glimpse ?
Paul
Le 9 mai 2012 à 13:05, Paul Libbrecht a écrit :
Hello XWiki users,
I'd like to know what the users of XWiki use as monitoring tools.
Does anyone have experience with Nagios? Zabbix? AlertSite? Another? A hand-made solution? (I haven't met one that is java based, that'd be cute!)
Here's my answer: - with i2geo.net, we used Nagios, it was relatively faithful at warning of downtime and even at warning "failed restart" (which remain a mystery to me but happen, from time to time)
- with curriki.org, we use AlertSite, it works very well to warn about the downtime and plot performance changes. Internally, also, we use a script to detect too long response times and restart (which happens when queries start to lock too many other queries) and a hand-made monitoring tool which displays a synchronized overview of the top, page loading time, apache and appservlogs, and mysql queries; it's a youngish but can help diagnose.
-- Sergiu Dumitriu http://purl.org/net/sergiu/