- observium - lists.observium.org

Poller count
by Jay Hanke 24 May '17

24 May '17

I'm frequently running out of memory on a 4 GB VM running CentOS7. The vm only runs Observium. I'm seeing a bunch of php /opt/observium/poller.php processes getting created and the box gradually runs out of memory and mariadb dies. I'm running version 8485. Any ideas? Thanks! Jay

3 2

function rrdtool_lastupdate
by yan flou 24 May '17

24 May '17

hi all, i have a question, can you please tell me what function rrdtool_lastupdate in (rrdtool.inc.php) makes, because i ty to integrate a new backend "opentsdb" for observium for the moment i retrive the data and the metric by rrdtool_update and i send him in to opentsdb but i need also the tags because the index aren't clear for example this is what i have for mempools put net.mempool 1495461482 404979712.000000 hostname=localhost ifIndex=used-host-resources-mib-6 put net.mempool 1495461482 7845933056.000000 hostname=localhost ifIndex=free-host-resources-mib-6 put net.mempool 1495461482 2097152.000000 hostname=localhost ifIndex=used-host-resources-mib-10 put net.mempool 1495461482 1071636480.000000 hostname=localhost ifIndex=free-host-resources-mib-10 put net.mempool 1495461542 8063819776.000000 hostname=localhost ifIndex=used-host-resources-mib-3 put net.mempool 1495461542 1260826624.000000 hostname=localhost ifIndex=free-host-resources-mib-3 but me i wante to have this put net.mempool 1495468562 8075042816.000000 hostname=localhost ifIndex=used-Virtual-memory put net.mempool 1495468562 1249603584.000000 hostname=localhost ifIndex=free-Virtual-memory put net.mempool 1495468562 410640384.000000 hostname=localhost ifIndex=used-Memory-buffers put net.mempool 1495468562 7840272384.000000 hostname=localhost ifIndex=free-Memory-buffers put net.mempool 1495468562 2097152.000000 hostname=localhost ifIndex=used-Swap-space put net.mempool 1495468562 1071636480.000000 hostname=localhost ifIndex=free-Swap-space can you help me please. ________________________________ yan flou

2 1

Re: [Observium] Exclusion in "poller.php -h all"
by davide.magni＠sorint.it 23 May '17

23 May '17

Thank you for your help. Is there a way to have a more frequent polling or is not possible at all? *Davide Magni* Il 05/22/2017 10:59 PM, Tom Laermans ha scritto: > Hi Davide, > > Sorry to have mailed you in private, just saw it, meant to go to the > list... > > Yes, you won't have more accurate graphs when you run a poller every > minute, RRD will only store a single sample every 5 minutes. I thought > we had some docs on that, but I can't seem to find them right now :o > > Tom > > On 22/05/2017 14:19, davide.magni(a)sorint.it wrote: >> >> Yes I think that when the CPU is too high on the appliances they go >> in the state of "down", but if it's a problem to use 10 minutes >> interval, i'm going to use 5. >> >> I'm using 1 minute interval for some other appliances because i need >> more accurate graphs, is that a problem too? >> >> Thank you >> >> *Davide Magni* >> >> Il 05/22/2017 12:36 PM, Tom Laermans ha scritto: >>> >>> Hi Davide, >>> >>> This is currently not possible. >>> >>> Why do you need this specific cron job? Are the appliances too slow >>> to finish within 5 minutes? >>> >>> Things like our RRD storage are geared towards 5 min polling, >>> running the job every 10 minutes will give you other issues instead... >>> >>> Tom >>> >>> >>> On 05/22/2017 11:31 AM, Davide Magni wrote: >>>> >>>> Hi, >>>> I have a question about the "poller.php -h all" added in crontab, >>>> to run every 5 minutes. >>>> >>>> I need to install a cronjob for son specific appliances, to run >>>> every 10 minutes: >>>> >>>>> */10 * * * * /app/observium/poller.php -h 48 >>>>> */10 * * * * /app/observium/poller.php -h 49 >>>>> */10 * * * * /app/observium/poller.php -h 50 >>>> Is it possible to exclude this specific appliances from the "-h >>>> all" option? >>>> >>>> Thank you. >>>> >>>> -- >>>> *Davide Magni* >>>> >>>> >>>> _______________________________________________ >>>> observium mailing list >>>> observium(a)observium.org >>>> http://postman.memetic.org/cgi-bin/mailman/listinfo/observium >>> >> >

2 1

Exclusion in "poller.php -h all"
by Davide Magni 22 May '17

22 May '17

Hi, I have a question about the "poller.php -h all" added in crontab, to run every 5 minutes. I need to install a cronjob for son specific appliances, to run every 10 minutes: > */10 * * * * /app/observium/poller.php -h 48 > */10 * * * * /app/observium/poller.php -h 49 > */10 * * * * /app/observium/poller.php -h 50 Is it possible to exclude this specific appliances from the "-h all" option? Thank you. -- *Davide Magni*

1 0

RealTime traffic view problem
by Richard Wigzell 19 May '17

19 May '17

Hello, we are running 17.5.8545 rolling and the Real time traffic graph is not working with the following error: "Can't get data about port (port id)" regardless of port id. This was previously working for all ports and now is not. [cid:image001.png@01D2D0A1.09D5BC00] We are also seeing the following errors in db.log when loading the page with the graph. [2017/05/19 12:54:44 +0100] index.php(3222): Failed dbQuery (#1054 - Unknown column 'pagpGroupIfIndex' in 'where clause'), Query: SELECT COUNT(*) FROM `ports` WHERE `pagpGroupIfIndex` = '10114' and `device_id` = '12' [2017/05/19 12:56:44 +0100] index.php(3222): Failed dbQuery (#1054 - Unknown column 'pagpGroupIfIndex' in 'where clause'), Query: SELECT COUNT(*) FROM `ports` WHERE `pagpGroupIfIndex` = '10114' and `device_id` = '12' [2017/05/19 12:57:44 +0100] index.php(16550): Failed dbQuery (#1054 - Unknown column 'pagpGroupIfIndex' in 'field list'), Query: SELECT `pagpGroupIfIndex` FROM `ports` WHERE `device_id` = '25' GROUP BY `pagpGroupIfIndex` [2017/05/19 12:57:58 +0100] index.php(3222): Failed dbQuery (#1054 - Unknown column 'pagpGroupIfIndex' in 'where clause'), Query: SELECT COUNT(*) FROM `ports` WHERE `pagpGroupIfIndex` = '10114' and `device_id` = '12' [2017/05/19 13:00:42 +0100] index.php(9137): Failed dbQuery (#1054 - Unknown column 'pagpGroupIfIndex' in 'where clause'), Query: SELECT COUNT(*) FROM `ports` WHERE `pagpGroupIfIndex` = '10114' and `device_id` = '12' [2017/05/19 13:02:56 +0100] index.php(16122): Failed dbQuery (#1054 - Unknown column 'pagpGroupIfIndex' in 'field list'), Query: SELECT `pagpGroupIfIndex` FROM `ports` WHERE `device_id` = '25' GROUP BY `pagpGroupIfIndex` Some help with fault finding this would be appreciated. Regards [cid:image001.jpg@01CE9215.D39244B0]<https://www.it.ox.ac.uk/> Richard Wigzell Networks, Security and Server Specialist IT Services | University of Oxford 13 Banbury Road, Oxford OX2 6NN T: 01865 616282 www.it.ox.ac.uk<https://www.it.ox.ac.uk/>

1 0

[RFE] - Ability to sort BGP sessions per type / uptime / updates
by Youssef BENGELLOUN - ZAHR 18 May '17

18 May '17

1 1

Temperature sensors
by Justin Miller 15 May '17

15 May '17

I’ve got a lot of equipment outside and as the weather flips from cool to hot to cool to hot I get a lot of alerts. I’ve got swings from 30 to 55 c. Is there a way to widen the auto generated limits across all system temperature oids? I tried adjusting the alert test conditions using arithmetic but I do not think it’s implemented. I tried this: sensor_value greater (@sensor_limit*1.25) sensor_value less (@sensor_limit_low*0.75) Justin

1 0

Changing IP scheme
by Jason Vanlandingham 14 May '17

14 May '17

I'm about to change the IP addresses of the Observium server and all network devices. Do I need to do anything other than change the hosts file? (Not running DNS)

2 1

95th percentile line wrong
by Tony Lill 13 May '17

13 May '17

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I'm using CE 0.16.10.8128, and I've noticed the the 95th percentile line is way too high on my graphs. As in, higher than the max value. In perusing the code, I noticed that it was based on in and out BITS, whereas everything else is based on OCTETS. Changing the definition to use inoctets and outoctets fixes this. - -- Tony Lill, OCT, ajlill(a)AJLC.Waterloo.ON.CA President, A. J. Lill Consultants (519) 650 0660 539 Grand Valley Dr., Cambridge, Ont. N3H 2S2 (519) 241 2461 - --------------- http://www.ajlc.waterloo.on.ca/ ---------------- -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iEYEARECAAYFAlkXUUoACgkQGS8yZq1uvxDHRwCbBQ41zE1medgwEpVV8tc+Vq9Y ka0An3eGRjD9t3scAmBm4I7vdZZJmiJH =eKWF -----END PGP SIGNATURE-----

1 0

Stuck alerts for sensors and statuses
by Robert Williams 12 May '17

12 May '17

Hi, Following a series of device upgrades and reloads over the last couple of days; we now have some 'invisible' stuck alerts: I cannot seem to find them anywhere, clicking them produces a blank list and even if you click the "Total" column. In this example I clicked the number "605" under Statuses but I get a page which only shows 595 items - so 10 missing, which equals the number of phantom alerts exactly. We had this once before and it was to do with devices, which are set to 'Ignore' for their alerts. We have 42 devices like this (a mixture of Lab and workstations) which unfortunately, would be a very large amount of manual checking. So I'm hoping there is a better way and I'm just missing it? :) So - any idea how I can find/reset/clear these items? Cheers! Robert Williams Custodian Data Centre Email: Robert(a)CustodianDC.com http://www.CustodianDC.com

1 0