- observium - lists.observium.org

Please help create an alert :) Nexus FEX PS problem
by Laszlo Nemeth 11 Aug '16

11 Aug '16

Hello At the night the Observium showed an alert (fortunately) that I never set up in the Alert checks . I see this alert in Global/Alerts and the left up box there is an “1 alert” in the Statuses row. If I click the “1 alert” it shows me that a Cisco Nexus FEX modul’s powersupply EVENT is alert and Status is offDenied. So it is a dead PS. What I can to do if I would like to get alert if this happen again? I have lot of Alert Checks but I am not able to do an alert for this PS check because I don’t know what is the correct Entity match of it. I tried the “sensor_class equals powersupply” and Test is “sensor_value greater 2” (looks the normal PS value is 2, now this failed PS value is 4). But it doesn’t work (no matched entities). Thanks Laszlo

2 2

alerting problem
by Laszlo Nemeth 11 Aug '16

11 Aug '16

Hello, At the morning I was on the 7933 but 2 hours ago I updated to 8039. I used a CPU alert for my routers. In the older version it is worked, I got the alerts via mail and PagerDuty as I set up. Now after the upgrade I would like to test the alerting because on 7933 i “pached” the pagerduty.inc.php for more info in the PD message (the default file just send like “something wrong somewhere” :) So. After the upgrade with the original (nor pached) PD file I would like a test to see how looks out a PD message now. I set a lower value for the CPU alerting (from 50% to 15%) to generate alerting (now the cpu load is 18-25%) After a few minutes I see in the Alers / Processor that my deviced appear and show that the status is FAILED. It is ok. In the Alert Logging I get the “Checks failed” so it is great. But. I never get “Alert notification sent” messages. (I get “Alert notification sent” for other events, so the alerting is working for other services). I see in the Alerts that the router’s status is Failed, Checked 18s Changed 20m 36s, Alerted: never. What I have to do for send the alerts? Or for the debug? Thanks BR Laszlo

3 7

Re: [Observium] Alerting on Application attributes
by Neil French 09 Aug '16

09 Aug '16

Hi, Thanks for the confirmation, it’s great that alerting on an OID will be coming soon and in my case that will work, I’ll keep the App for viewing graphs and use a custom OID to monitor alerts by. Ultimately alerting on data collected by apps would be beneficial but not essential, as it will give neater integration options but I can see the complexities in that, maybe allowing apps to store pertinent stats for alerting in a column holding an array with a sub-array for each app? Anyway, I look forward to seeing the OID alerting soon. Thanks, Neil. ------------------------------ Message: 9 Date: Thu, 04 Aug 2016 09:50:00 -0500 From: Adam Armstrong <adama(a)memetic.org <mailto:adama@memetic.org>> To: Observium Network Observation System <observium(a)observium.org <mailto:observium@observium.org>> Cc: observium(a)observium.org <mailto:observium@observium.org> Subject: Re: [Observium] Alerting on Application attributes Message-ID: <fd070f1f-899c-4772-896b-540b771c7c16(a)typeapp.com <mailto:fd070f1f-899c-4772-896b-540b771c7c16@typeapp.com>> Content-Type: text/plain; charset="utf-8" Hi, Alerting on custom oid is planned for the not too distant future (along with some other additions to custom oid). Alerting on apps is much more complex, I'm not entirely sure how to approach it at the moment. I suspect it may not look exactly like the current alerting, but it is probably some time away. Thanks, Adam. Sent from BlueMail On 4 Aug 2016, 09:34, at 09:34, Neil French <neil(a)numbergroup.com <mailto:neil@numbergroup.com>> wrote: > Hi, > > Apologies if this has been covered previously, I can’t seem to browse > any mailing list archives as the link is dead. > > I have previously used custom OIDs to monitor a VoIP Application > (FreeSWITCH) but found I couldn’t alert on custom OIDs, and that the > graphing wasn’t as customisable as I wanted, so I have built a custom > application for freeswitch. > > I now want to be able to raise alerts on a few things, such as when the > number of new sessions per second is over a predefined value or when > the uptime resets (application pop and restart). > > I can’t seem to find a way of creating an alert check for anything > that’s not one of the basic properties of a device though. > > Could someone please tell me if this is possible? For example can I > create my own attributes in the device then query them with an alert > check? I have tried adding something like: > > $device[‘device_state’].[’my parameter’] = value > > But that doesn’t seem to be stored outside of the php script. > > It’s quite important that we are able to alert on custom parameters and > I can either provide these as OIDs or through the unix agent, whichever > gets the job done. > > > > Thanks, > > > > > Neil. > _______________________________________________ > observium mailing list > observium(a)observium.org <mailto:observium@observium.org> > http://postman.memetic.org/cgi-bin/mailman/listinfo/observium <http://postman.memetic.org/cgi-bin/mailman/listinfo/observium>

1 0

OpenWRT && wifi
by Ciro Iriarte 08 Aug '16

08 Aug '16

Hi!, anybody knows if it's possible to take wifi performance data from OpenWRT?. Regards, CI.-

2 1

DeltaUPS graphs broken after r8016
by Kent Johannessen 08 Aug '16

08 Aug '16

Well not really... but it will not use renamed RRDs before after a new discovery of sensors. Meanwhile polling will create new RRDs with old names, so it seems you have lost all data. Several hours of data can be missed because of this. Maybe some improvements can be done for that. NOTICE: Please immediately e-mail back to sender if you are not the intended recipient. Thereafter delete the e-mail along with any attachments without making copies. The sender reserves all rights of privilege, confidentiality and copyright.

3 5

Holes in graphs
by Zoellers, Billy 08 Aug '16

08 Aug '16

Hello all, We're seeing holes on our graphs (example attached). This appears to be affecting all graphs in our installation on all devices. I know our connection is good (affecting LAN and WAN devices), but looking for any other ideas for correcting this? We have also tried increasing the number of poller processes. Only around 1000 ports on this install, 8 GB of RAM and 4 vCPU. Thanks, Billy [cid:image001.png@01D1EEFD.85030800] [cid:DeanDortonbox.png] Billy Zoellers Infrastructure Engineer 106 W. Vine St., Ste 600 Lexington, KY 40507 ph (859) 425-7640 fax (859) 425-3640 CONFIDENTIALITY NOTICE: The preceding email message, and any files or attachments transmitted with it, may contain confidential and privileged information, and may constitute protected work product, confidential client information, trade secrets, and/or other non-public information. Furthermore, this message and any attachments may be protected by accountant/client privilege, attorney/client privilege, and/or other applicable privileges. Any unauthorized review, use, dissemination, distribution, disclosure, or reproduction of this message is strictly prohibited and may be unlawful. This message is intended to be conveyed and transmitted only to the designated recipient(s). If you are not an intended recipient of this message, please notify the sender immediately by email or by telephone at 859-255-2341. Also, please delete this email message including all attachments and destroy all copies.

4 7

VictorOps and Observium
by Andy Lemin 08 Aug '16

08 Aug '16

Hi guys, Can I please add a +11 to this one please ;) http://jira.observium.org/browse/OBSERVIUM-1800 VictorOps is great, and alerting through Slack is just not good enough for critical incidents. Thanks, Andy.

2 2

Graphs no longer draw after restore
by Vlad Kratsberg 08 Aug '16

08 Aug '16

Hi, I recently restored observium to a different machine. I restored the following: config.php mysql db rrd Everything looks good except for the fact that graphs stopped drawing and display data in a real time. When i click on port to see real time data, i get "can't get data about port xxx". See attached images. Any ideas ? Thanks

3 4

No memory pools detected for OSX hosts
by Oliver Baker 08 Aug '16

08 Aug '16

Hi there, I'm running Observium 0.16.7.8003 (stable) for monitoring OSX hosts which is working great, except no memory pools are being detected for any hosts apart from "Swap space". Memory for OSX hosts did work previously, but after I mailed in re: "Incorrect free physical memory for Unix hosts" and Adam A. made a change, the memory pools have disappeared. What's the best course of action? Many thanks Olly

1 0

Apps tab used to show a single graph per item.
by Ricardo Meleschi 07 Aug '16

07 Aug '16

Hello Observium overlords, In the past the Apps tab in observium would only show a single graph per item/array, but with the later versions it shows four graphs per item of varying timelines. Is it possible to configure Observium to only show a single timeline graph per array? When a large number of servers are running the application, it’s difficult to compare graphs when they are displayed in the current manner. Thanks, Ricardo

2 1