No. To keep track of this would cause issues for performance, since it'd require four extra fields in the ports table.
Adam.
Sent from BlueMail
On 27 Jun 2016, 15:31, at 15:31, "Moerman, Maarten" mmoerman@ebay.com wrote:
Mr Armstrong? Is this possible?
From: observium <observium-bounces@observium.orgmailto:observium-bounces@observium.org> on behalf of prost pierrick <pierrick.prost@cnrs.frmailto:pierrick.prost@cnrs.fr> Reply-To: Observium Network Observation System <observium@observium.orgmailto:observium@observium.org> Date: Monday, June 27, 2016 at 4:29 PM To: Observium Network Observation System <observium@observium.orgmailto:observium@observium.org> Subject: Re: [Observium] Delay the alert trigger in case of loss only one ping
Hi,
Thanks everyone for your responses …after investigated, I discovered a lot a packets Discard (and no packetError) on my openVswitch (promox VE), luckily I monitor my proxmox hypervisor (Debian) over Observium ☺
[cid:image001.png@01D1D091.0131EEE0]
Maarten, I try to reproduce your alert checker exemple but with Discards packets … but no metrics for discard ?
https://www.observium.org/docs/attribs_metrics/
Pierrick
De : observium [mailto:observium-bounces@observium.org] De la part de Moerman, Maarten Envoyé : lundi 27 juin 2016 11:40 À : Observium Network Observation System Objet : Re: [Observium] Delay the alert trigger in case of loss only one ping
Hi, my 2 cents:
I have a delay of 1 on my device_down alert, just to prevent false positives. SNMP just sometimes has a hickup. Although usually I don’t have these, but in order to prevent panic/heart attacks, I did introduce that 1 time delay on my device down alerts.
Haven you applied alert checkers on port-errors all around your network as well? CRC/port errors, and things like these you might want to fix first before you actually get in to device down alerting. Just a simple "ifInErrors_rate gt 0 & ifOutErrors_rate gt 0 on ifType equals ethernetCsmacd” alert-check will give you all these ports. you’d be surprised how much stuff is wrong in an average network. Also certainly check the port to which your observium machine is connected to, and make sure to monitor your observium machine as well, make sure your machine is well specced out.
In addition to this, I’ve configured IP SLA on all our core routers to see/verify that latency on our links is within the SLA parameters of our suppliers.
Do this before you check on devices down.
Maarten
Maarten Moerman | Mgr, Network Engineering | eBay Classifieds | +31-655122247 | mmoerman@ebay.commailto:mmoerman@ebay.com
From: observium <observium-bounces@observium.orgmailto:observium-bounces@observium.org> on behalf of Adam Armstrong <adama@memetic.orgmailto:adama@memetic.org> Reply-To: Observium Network Observation System <observium@observium.orgmailto:observium@observium.org> Date: Monday, June 27, 2016 at 11:16 AM To: Observium Network Observation System <observium@observium.orgmailto:observium@observium.org> Subject: Re: [Observium] Delay the alert trigger in case of loss only one ping
Delay is in poller intervals.
If you're getting notifications that your devices are down, you have some pretty serious network problems, as it takes quite a lot to actually mark something as down.
Adam.
Sent from BlueMailhttp://www.bluemail.me/r On 27 Jun 2016, at 10:07, prost pierrick <pierrick.prost@cnrs.frmailto:pierrick.prost@cnrs.fr> wrote: Hello,
I have make some tries but I’m still spammed with fake alerting.
Delay value is in minute ? How can I debug poller to check if I really lost ping ? (I’m on a 10 gbit backbone network.. I’m really surprise).
Regards
pierrick
De : prost pierrick Envoyé : jeudi 23 juin 2016 12:36 À : Observium Network Observation System Objet : RE: [Observium] Delay the alert trigger in case of loss only one ping
Delay ..yes i see it !
Thanks have a good day.
De : observium [mailto:observium-bounces@observium.org] De la part de Adam Armstrong Envoyé : jeudi 23 juin 2016 12:21 À : Observium Network Observation System Cc : observium@observium.orgmailto:observium@observium.org Objet : Re: [Observium] Delay the alert trigger in case of loss only one ping
it isn't just one ping, it tries multiple before it marks a device as down.
You can set the delay field, which will cause the alert to wait for x number of down periods before it triggers.
Adam.
Sent from BlueMailhttp://www.bluemail.me/r On 23 Jun 2016, at 10:39, prost pierrick <pierrick.prost@cnrs.frmailto:pierrick.prost@cnrs.fr> wrote: Hi everyone , Observium lost 1 ping sometime (I don’t know why ) after each ping lost, alerts are triggered.
Actually I have only two Alerts Checks :
Device Down device_status equals 0 * * Device UP device_status equals 1 *
It’s possible to add a delay before triggering ? Like waiting 5 ping lost before triggering ?
Regards.
Pierrick Prost
observium mailing list observium@observium.orgmailto:observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.orgmailto:observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium