Alert Checker for memory?
I created an alert to let me know when a host uses over 90% of its memory. The problem is that linux uses over that in "cached" memory. Is there a way to exclude cached memory? Physical and Swap are fine.
I am using mempool_perc greater 90 in Check Conditions.
Thanks, ----------------------- Leo Dano
You can exclude entities by mempool_descr, perhaps "mempool_descr notmatch *Cache*" ?
adam.
Sent from Mailbird [http://www.getmailbird.com/?utm_source=Mailbird&utm_medium=email&utm...] On 23/06/2016 22:04:59, Dano, Leo ldano@qualcomm.com wrote: I created an alert to let me know when a host uses over 90% of its memory. The problem is that linux uses over that in “cached” memory. Is there a way to exclude cached memory? Physical and Swap are fine. I am using mempool_perc greater 90 in Check Conditions. Thanks, ----------------------- Leo Dano
Actually spoke too soon and did it with entity match
Associations Add Device Match
Entity Match
*
mempool_descr match *Physical*
*
mempool_descr match *Swap*
*
mempool_descr match *Virtual*
Thanks, ----------------------- Leo Dano
From: observium [mailto:observium-bounces@observium.org] On Behalf Of Adam Armstrong Sent: Thursday, June 23, 2016 2:25 PM To: observium@observium.org Subject: Re: [Observium] Alert Checker for memory?
You can exclude entities by mempool_descr, perhaps "mempool_descr notmatch *Cache*" ?
adam.
Sent from Mailbirdhttp://www.getmailbird.com/?utm_source=Mailbird&utm_medium=email&utm_campaign=sent-from-mailbird
On 23/06/2016 22:04:59, Dano, Leo <ldano@qualcomm.commailto:ldano@qualcomm.com> wrote: I created an alert to let me know when a host uses over 90% of its memory. The problem is that linux uses over that in “cached” memory. Is there a way to exclude cached memory? Physical and Swap are fine.
I am using mempool_perc greater 90 in Check Conditions.
Thanks, ----------------------- Leo Dano
It'd be cleaner and slightly quicker to exclude bad things in a single match entry, otherwise it's taking three queries instead of one :)
adam.
Sent from Mailbird [http://www.getmailbird.com/?utm_source=Mailbird&utm_medium=email&utm...] On 23/06/2016 22:28:17, Dano, Leo ldano@qualcomm.com wrote: Actually spoke too soon and did it with entity match Associations Add Device Match Entity Match * mempool_descr match *Physical* * mempool_descr match *Swap* * mempool_descr match *Virtual* Thanks, ----------------------- Leo Dano
From: observium [mailto:observium-bounces@observium.org] On Behalf Of Adam Armstrong Sent: Thursday, June 23, 2016 2:25 PM To: observium@observium.org Subject: Re: [Observium] Alert Checker for memory? You can exclude entities by mempool_descr, perhaps "mempool_descr notmatch *Cache*" ? adam. Sent from Mailbird [http://www.getmailbird.com/?utm_source=Mailbird&utm_medium=email&utm...] On 23/06/2016 22:04:59, Dano, Leo <ldano@qualcomm.com [mailto:ldano@qualcomm.com]> wrote: I created an alert to let me know when a host uses over 90% of its memory. The problem is that linux uses over that in “cached” memory. Is there a way to exclude cached memory? Physical and Swap are fine. I am using mempool_perc greater 90 in Check Conditions. Thanks, ----------------------- Leo Dano
That's only on checker association rebuild though, not poller. So yes, better, but no impact on polling speed, which is most important.
On 23/06/2016 23:29, Adam Armstrong wrote:
It'd be cleaner and slightly quicker to exclude bad things in a single match entry, otherwise it's taking three queries instead of one :)
adam.
Sent from Mailbird http://www.getmailbird.com/?utm_source=Mailbird&utm_medium=email&utm_campaign=sent-from-mailbird
On 23/06/2016 22:28:17, Dano, Leo ldano@qualcomm.com wrote:
Actually spoke too soon and did it with entity match
It does affect discovery, and the rebuilding can get pretty slow on very large installs. Perhaps not with memory stuff, but this sort of thing is significant for ports or sensors!
adam.
Sent from Mailbird [http://www.getmailbird.com/?utm_source=Mailbird&utm_medium=email&utm...] On 23/06/2016 22:31:36, Tom Laermans tom.laermans@powersource.cx wrote: That's only on checker association rebuild though, not poller. So yes, better, but no impact on polling speed, which is most important.
On 23/06/2016 23:29, Adam Armstrong wrote:
It'd be cleaner and slightly quicker to exclude bad things in a single match entry, otherwise it's taking three queries instead of one :)
adam.
Sent from Mailbird [http://www.getmailbird.com/?utm_source=Mailbird&utm_medium=email&utm...] On 23/06/2016 22:28:17, Dano, Leo ldano@qualcomm.com [mailto:ldano@qualcomm.com] wrote: Actually spoke too soon and did it with entity match
Ah right, forgot discovery rechecks those. Yeah, it all adds up obviously.
Tom
On 23/06/2016 23:32, Adam Armstrong wrote:
It does affect discovery, and the rebuilding can get pretty slow on very large installs. Perhaps not with memory stuff, but this sort of thing is significant for ports or sensors!
adam.
Sent from Mailbird http://www.getmailbird.com/?utm_source=Mailbird&utm_medium=email&utm_campaign=sent-from-mailbird
On 23/06/2016 22:31:36, Tom Laermans tom.laermans@powersource.cx wrote:
That's only on checker association rebuild though, not poller. So yes, better, but no impact on polling speed, which is most important.
On 23/06/2016 23:29, Adam Armstrong wrote:
It'd be cleaner and slightly quicker to exclude bad things in a single match entry, otherwise it's taking three queries instead of one :)
adam.
Sent from Mailbird http://www.getmailbird.com/?utm_source=Mailbird&utm_medium=email&utm_campaign=sent-from-mailbird
On 23/06/2016 22:28:17, Dano, Leo ldano@qualcomm.com wrote:
Actually spoke too soon and did it with entity match
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
participants (3)
-
Adam Armstrong -
Dano, Leo -
Tom Laermans