Re: Alerting query

[Martin Pala]

Hi,

currently it is not possible to group the alerts like this. We will extend monit configuration language to allow for complex test conditions which will solve this problem. We have no schedule for the implementation yet, the priority is graphs/trends implementation for M/Monit 3.0 now, we'll start implementing Monit conditions after it.

Best regards,

Martin

On Sep 15, 2010, at 11:07 AM, Andy Scutt wrote:

Hi,

 

I was looking through the docs to see if there was any way to limit alerting but couldn't find alerting but couldn't find anything, hopefully someone out there has had the same issue and can educate me :)

 

We monitor many systems for clients, each customer system is setup as a host in monit. Each host has multiple checks (tcp port and url checks). We'd like to limit alerts so that if more than one check fails for a host we only get one set of alerts not alertas for each failed check, is this possible?

 

For example, heres a sample client config:

 

check host Customer_A with address client.address.com

#http check
if failed port 80 proto http timeout 20 seconds then alert

#https check
if failed port 443 type TCPSSL proto http timeout 20 seconds then alert
#node manager check

if failed host 192.168.40.73 port 9400 for 2 cycles then alert
if failed host 192.168.40.76 port 9400 for 2 cycles then alert
#web app check

if failed url http://192.168.40.73:8006/activerecruiter/companies/monitor/jsp-container-monitor.jsp and content == 'JSPContainerOk' timeout 20 seconds for 2 cycles then alert
if failed url http://192.168.40.76:8006/activerecruiter/companies/monitor/jsp-container-monitor.jsp and content == 'JSPContainerOk' timeout 20 seconds for 2 cycles then alert
if failed url http://192.168.40.73:8006/activerecruiter/companies/monitor/db-connection-monitor.jsp and content == 'DatabaseOk' timeout 20 seconds for 2 cycles then alert
if failed url http://192.168.40.76:8006/activerecruiter/companies/monitor/db-connection-monitor.jsp and content == 'DatabaseOk' timeout 20 seconds for 2 cycles then alert
#admin check
if failed url http://192.168.40.73:6900/monitor/admin_test.jsp and content == 'Admin' timeout 20 seconds for 2 cycles then alert

Worst case if the whole client system dies we get 9 alerts per cycle, when one would do. is this possible?

 

Many thanks

Andy

 

 

--
To unsubscribe:
http://lists.nongnu.org/mailman/listinfo/monit-general