- observium - lists.observium.org

Reduce your storage IO with rrdcached
by Markus Klock 27 Oct '14

27 Oct '14

There was some discussion on IRC the other day about rrdcached and how it can be used with Observium to reduce your storage IO a lot when polling a huge amount of interfaces. However, not a lot of documentation exists yet on how to get rrdcached to work with observium. I did some trial and error and found a working method and wrote a quick and dirty guide. You can find it here: http://blog.best-practice.se/2014/10/using-rrdcached-with-observium.html Feedback is highly appreciated! /Markus

5 7

Re: [Observium] Several broken graphs
by Adam Armstrong 26 Oct '14

26 Oct '14

1 0

Several broken graphs
by Robert Williams 26 Oct '14

26 Oct '14

Hi, A number of our hosts have suddenly got broken graphs for items like: * Ping Response * SNMP Response * Poller Duration * Running Processes * Users Logged In * Context Switches * System Interrupts When in debug mode, the graph is showing these errors (depending which graph): command returned 1 (ERROR: String ends after the = sign on 'CDEF:ping_min=' ) command returned 1 (ERROR: String ends after the = sign on 'CDEF:val_min=' ) command returned 1 (ERROR: String ends after the = sign on 'CDEF:value_min=' ) command returned 1 (ERROR: String ends after the = sign on 'CDEF:users_min=' ) command returned 1 (ERROR: String ends after the = sign on 'CDEF:procs_min=' ) An example of a full string for a broken graph is: graph /tmp/0H1FmUUbABhyOl7U.png --alt-autoscale --rigid --start 1414240174 --end 1414326574 --width 1159 --height 300 -c BACK#EEEEEE00 -c SHADEA#EEEEEE00 -c SHADEB#EEEEEE00 -c FONT#000000 -c CANVAS#FFFFFF00 -c GRID#a5a5a5 -c MGRID#FF9999 -c FRAME#5e5e5e -c ARROW#5e5e5e -R normal --font LEGEND:8:'DejaVuSansMono' --font AXIS:7:'DejaVuSansMono' --font-render-mode normal -E COMMENT:' ' COMMENT:' Now' COMMENT:' Avg' COMMENT:' Min' COMMENT:' Max' COMMENT:' Total' COMMENT:'\l' DEF:procs=/opt/observium/rrd/hostname.was.here/hr_processes.rrd:procs:AVERAGE CDEF:procs_min= CDEF:procs_max= AREA:procs#E881B1:"Processes " LINE1.5:procs#744058 GPRINT:procs:LAST:%6.1lf%S GPRINT:procs:AVERAGE:%6.1lf%S GPRINT:procs_min:MIN:%6.1lf%S GPRINT:procs_max:MAX:%6.1lf%S COMMENT:'\l' Anyone else seeing this or has our install spontaneously broken itself? We are at 5910 on Ubuntu 14.04.1 - cheers! Robert Williams Custodian Data Centre Email: Robert(a)CustodianDC.com http://www.CustodianDC.com

1 0

Ports not found on older PowerConnect switches
by Dan Ratzlaff 25 Oct '14

25 Oct '14

Hello. I'm having trouble getting Observium (CE 0.14.4.5229) to recognize ports on my Dell PowerConnect 3048 and 3024 switches. It appears these switches only support SNMP v1. Observium connects and grabs SNMP contact, location, etc. but doesn't see ports. Here's output from add and discovery for my 3024 switch: ----- # ./add_device.php switch-ne Observium v0.14.4.5229 Add Device(s) Try to add switch-ne: Trying v2c community public ... No reply on community public using v2c. Could not reach switch-ne with given SNMP community using v2c. Trying v1 community public ... Now discovering switch-ne (id = 29) switch-ne 29 powerconnect Ports : XXXXXXXXXXXXXXXXXXXXXXXXXX Discovered in 0.498 seconds switch-ne 29 powerconnect IPv4 Addresses : Discovered in 0.092 seconds switch-ne 29 powerconnect IPv6 Addresses : Discovered in 0.189 seconds Added device switch-ne (29). Devices added: 1. ----- # ./discovery.php -h 29 Observium v0.14.4.5229 Discovery switch-ne 29 powerconnect OLD: powerconnect, NEW: powerconnect --- Ports : XXXXXXXXXXXXXXXXXXXXXXXXXX Port Stacks: Processors : Dell-Vendor-MIB hrDevice: Memory : Dell-Vendor-MIB HOST-RESOURCES-MIB IPv4 Addresses : IPv6 Addresses : Sensors: FASTPATH-BOXSERVICES-PRIVATE-MIB + Dell-Vendor-MIB Storage : Inventory: Physical Inventory : hrDevice : Discovery protocols: LLDP-MIB CISCO-CDP-MIB OSPF Neighbours: ARP/NDP Tables : BGP Sessions: No BGP on host VLANs: Q-BRIDGE-MIB UCD Disk IO : Discovered in 2.012 seconds Current Revision : 5229 New Revision : 5909 MySQL: Cell[2/0s] Row[19/0.01s] Rows[33/0.02s] Column[0/0s] Update[1/0s] Insert[5/0.02s] Delete[0/0s] ----- I can do a snmpwalk on these switches from the same machine that Observium is running on and can see all the interface information. Is there something I need to tweak in Observium to get these switches discovered correctly? Thanks! -- Dan

2 5

allow graphs to be viewed by anyone / config.php vs web configuration
by Massimo Sala 25 Oct '14

25 Oct '14

Hi ! Observium CE 0.14.4.5229 Following the docs, in config.php I set : $config['auth_mechanism'] = "mysql"; $config['allow_unauth_graphs'] = 1; $config['allow_unauth_graphs_cidr'] = array( "10.1.0.0/16" ); I create a user "guest" ... but it cannot see any graph. I login to Observium as admin, I edit the user "guest" properties : Device Access None Configured Grant access to new device Interface Access None Configured Grant access to new interface I see different configuration : config.php vs web configuration. Is there any mismatch ? How can I grant a full "read-only" access to a "guest" user ? Best regards, Sala

2 1

observium customizations on PHP code
by Massimo Sala 24 Oct '14

24 Oct '14

Observium CE 0.14.4.5229 PHP 5.4.4-14+deb7u14 MySQL 5.5.40-MariaDB-1~wheezy-log SNMP NET-SNMP version: 5.4.3 RRDtool 1.4.7 I have modified one source file : /opt/observium/includes/alerts.inc.php I restarted web server and php-fastcgi ( no opcode cache ), but Observium doesn't see the update. Is there an internal function caching ? Am I missing something ? Best regards, Sala

3 2

How can I unsubscribe?
by Franck Burgaudeau 24 Oct '14

24 Oct '14

Hi, I would unsubscribe from this miling list but when I enter my mail in this page, I don't receive anything and I continue to receive all mail... Thank you for your reply. Regards *Franck Burgaudeau* f.burgaudeau(a)graphique-alliance.com <http://www.graphique-alliance.com/> *Attention !!* *Graphique Alliance a déménagé.Les téléphones, fax et e-mails sont inchangés.* *Nouvelle Adresse : * *50 rue Laudinat - 33130 Bègles**Graphique**Alliance* Intégration, développement et formation pour les industries graphiques 50 rue Laudinat 33130 Bègles Standard : +33 (0)5 56 67 55 78 Téléphone : +33 (0)1 85 08 64 79 Portable : +33 (0)6 75 61 75 73 Fax : +33 (0)5 57 54 12 77

1 0

Re: [Observium] polling issues with 5897
by Rutger Bevaart 24 Oct '14

24 Oct '14

A strange issue indeed… Based on Mike’s input I ran on the old-style poller for a day, then all is fine. I’m now running on poller-wrapper for another day, with concurrency set to 2, and that seems to be fine as well. I had it at 8, then all goes to h*ll. Will try ‘4’ today ;-) Regards, Rutger > > Netscreen firewall > > [https://observium.exegy.net/graph.php?type=device_processor&device=3&to=141…] > > Arista Switch > [https://observium.exegy.net/graph.php?type=device_processor&device=108&to=1…] > > Cisco Router > [https://observium.exegy.net/graph.php?type=device_uptime&device=70&to=14140…] > > > I can send you a debug poller.php run privately if you'd like. >

1 0

Multiple tun0 interfaces showing up in graphs
by Gary Tang 24 Oct '14

24 Oct '14

For some reason, I am getting multiple tun0 interfaces in graphs, even though I only have one via openVPN. Where should I start looking to debug this? Graph: http://postimg.org/image/b876gqbq5/ Thanks for looking, Gary

1 0

Re: [Observium] Alerts in "Other" state
by Adam Armstrong 24 Oct '14

24 Oct '14

I have no idea. How long is a piece of string? Guys, I'm not here to tickle you until you actually give some information. Give as much information as you have about a problem or you won't get it fixed. Life is too short to chase you all for screen shots and more than 3 word descriptions of your problems. adam. ------ Original Message ------ From: "Ben Steele" <ben(a)bensteele.org> To: "Adam Armstrong" <adama(a)memetic.org>; "Observium Network Observation System" <observium(a)observium.org> Sent: 10/23/2014 4:02:08 PM Subject: Re: [Observium] Alerts in "Other" state >What might cause that? > >The alerts that haven't been checked aren't new and as previously >mentioned even something simple like is device up/down has these >non-checked alerts so it's not like the alert isn't applicable to the >device. > >I've checked a device that has a non-checked alert and its graphs and >polling look fine. > >Cheers > >On 24/10/2014 3:33 am, "Adam Armstrong" <adama(a)memetic.org> wrote: >>It means the alert has never been checked. >> >>adam. >> >>------ Original Message ------ >>From: "Ben Steele" <ben(a)bensteele.org> >>To: "Observium Network Observation System" <observium(a)observium.org> >>Sent: 10/22/2014 8:15:36 PM >>Subject: [Observium] Alerts in "Other" state >> >>>Hi, >>> >>>I have quite a few alerts stuck in the "Other" state, any idea on >>>what causes this? >>> >>>I had this problem a while ago and then it seemed to disappear after >>>creating a few more alerts and several svn updates so i'm not really >>>sure what cleared it but after adding more devices it's back again. >>>I've tried regenerating the alerts a few times but that makes no >>>difference, is there some debug that can be turned on to work out >>>what is going on here? >>> >>>It's across multiple alert checkers and even something as simple as >>>"device status eq 0" for a device down check has multiple alerts >>>stuck in "other" state. Hovering over the alert shows an empty >>>status: >>> >>> >>> >>> >>> >>>Cheers >> >>_______________________________________________ >>observium mailing list >>observium(a)observium.org >>http://postman.memetic.org/cgi-bin/mailman/listinfo/observium >>

2 1