- observium - lists.observium.org

Discover new devices problem
by zois roupas 10 Oct '16

10 Oct '16

Hello everyone, Great Documentation and so far it's a great tool. One question though because i can't find a way to get auto discover new devices to work. Everything works great when adding the device ./add_device.php <ip> public v2c but when running ./discovery.php -h new , 0 devices discovered. I guess it's not a problem with the SNMP cause it wouldn't work with the manual ip addition, right? So the next step was to check the hosts file from the observium server. so this is the config 192.168.1.2 test-01.home.com #(machine to discover) The sysName variable from the machine that i want to discover is Linux test-01 3.10.0-327.36.1.el7.x86_64 #1 SMP Sun Sep 18 13:04:29 UTC 2016 x86_64 x86_64 The firewall is disabled in both machines Sorry for the trouble but what am i missing here? Thanks for the time Zois

1 0

syslog integration issue
by Grégory Lemblé 10 Oct '16

10 Oct '16

Morning, I try to implement the syslog.php features inside our Observium instance. on debug mode, we got syslog info inside the logs/debug.log file. But on the web side, impossible to get any message from the syslog. I'm not sure which area i need to investigate to get it done. Some help will be appreciate. Many Thanks ----------- Observium Version Pro 0.16.10.8118 (Stable) ----------- rsyslogd 7.4.4, compiled with: FEATURE_REGEXP: Yes FEATURE_LARGEFILE: No GSSAPI Kerberos 5 support: Yes FEATURE_DEBUG (debug build, slow code): No 32bit Atomic operations supported: Yes 64bit Atomic operations supported: Yes Runtime Instrumentation (slow code): No uuid support: Yes ----------- /etc/rsyslog.d/60-observium.conf: #--------------------------------------------------------- #send remote logs to observium $template observium,"%fromhost%||%syslogfacility%||%syslogpriority%||% syslogseverity%||%syslogtag%||%$year%-%$month%-%$day% %timereported:8:25%||%msg%||%programname%\n" $ModLoad omprog $ActionOMProgBinary /var/www/observium/syslog.php :inputname, isequal, "imudp" :omprog:;observium #& ~ & stop #--------------------------------------------------------- ----------- debug.log file: [2016/10/10 08:00:15 +0800] syslog.php(19124): hostname||1||4||4||auth:||2016-10-10 08:00:14|| User XXXXXXXXXXXX logout||auth [2016/10/10 08:00:15 +0800] syslog.php(19124): hostname||1||6||6||Security:||2016-10-10 08:00:15|| SSH login by public key from src IP XXXXXXX from src MAC XXXXX to USER EXEC mode using DSA as Server Host Key. ||Security ----------- config.php: $config['syslog']['debug'] = True; $config['enable_syslog'] = 1; // Enable Syslog $config['syslog']['fifo'] = FALSE; ------------- Front Web: No syslog entries found!Check that the syslog daemon and Observium configuration options are set correctly, that your devices are configured to send syslog to Observium and that there are no firewalls blocking the messages. See documentation <http://www.observium.org/wiki/Category:Documentation> and configuration options <http://www.observium.org/wiki/Configuration_Options#Syslog_Settings> for more information

1 0

Re: [Observium] Auto discovery
by Mikey van der Worp 07 Oct '16

07 Oct '16

To whom it may concern, Currently we are having some issues with Observium's Auto Discovery. The problem is that it currently only discovers about 10% of the network. Does somebody have a clue on how to fix this issue? We already disabled the option "require valid hostnames" at the auto discovery settings. Unfortunately after changing this option, it only discovered about 10 more devices. Also we do not have any problems with the CDP neighbours on our switches. Best regards, Mikey van der Worp -- Mikey van der Worp System Engineer Utelisys Communications B.V. Trinity Buildings Tower A, 7th floor Pietersbergweg 15 1105BM Amsterdam M +31 (0) 62 942 2052 T +31 (0) 20 561 8010 F +31 (0) 20 561 8021 www.utelisys.com<http://www.utelisys.com> - https://www.utelisys.com/

3 2

Unable to save sensor values
by Robert Williams 07 Oct '16

07 Oct '16

Hi, On devices with large numbers of sensors it appears that I'm unable to set custom limits any more. I'm sure there used to be something about a maximum number of rows to be updated (a SQL setting?) but I can't remember or find any reference of it. The path to the page is: /device/device=XXX/tab=edit/section=sensors/ I don't get any errors when applying, but, as soon as the page refreshes the old settings are back. In fact, any change on the sensors page fails (setting custom values, disabling alerts, enabled custom limits etc.). It works fine on all devices with lower numbers of sensors. These ASR9k have well over 150 I reckon, maybe more, it's hard to count. Anyway, can someone remind me of the SQL tweak for this? Or is this a bug? Cheers! Robert Williams Custodian Data Centre Email: Robert(a)CustodianDC.com http://www.CustodianDC.com

2 2

Restrict alerts to HA active F5 only
by Daniel Horth 06 Oct '16

06 Oct '16

Hi all I have an alert checker set up to monitor our F5 BIG-IP devices, alerting administrators if they forget to syncronise config after making changes. My alert looks like this: Device Match: * Entity Match: status_type equals f5-config-sync-state Test conditions (Require all): status_name notequals inSync status_name notequals standalone This is working great, however to reduce spamminess of alerts I am looking to only send the alert for the primary (HA active) device; when an additional status_type / condition match is true, i.e.: Entity Match: status_type equals f5-ha-state Test conditions: status_name equals active Is it possible to have alert checkers trigger based upon groups of entity matches like this? Am I approaching this in the wrong way? Thanks Dan

1 0

Syslog Alerts
by Brad Bendy 05 Oct '16

05 Oct '16

Hi, First off, the last few revisions have been killer, it's getting better every few weeks it seems, thanks! Ive setup some syslog alerts and I see under "syslog alerts" it will show a entry and "notified" will be no, then it appears after a few minutes this changes to yes and send the notification. It looks like it does not send the message until it sees more then X number of messages, It's sent after I get this in the logs: 2016-10-04 20:53:13Error (3)LAST message repeated 10 times 2016-10-04 20:52:47Error (3)LAST message repeated 7 times 2016-10-04 20:52:33Error (3)XNTPD NTP Server Unreachable Is this documented somewhere and Im missing it? Just trying to see when the message gets fired off. We were using some janky omprog code to send alerts that were "instant" but rather have Observium send them as it's much cleaner all around, but some alerts we want to get on the very first message and right away as they are critical. I noticed the syslog rules severity of a 8, im guessing eventually that will control how "sensitive" the alerts go out maybe? Thanks

1 0

CBQoS Issue & Temperature
by Talk Jesus 05 Oct '16

05 Oct '16

Hello First thing, does the attach show anything I need to be concerned about? The warnings, although 6 days ago I'm not understanding where they came from. These are new switches, all 3 fans installed and green lights. Airflow sufficient and cooled. Secondly, I have 3 ports showing a CBQoS error. I checked and see it's due to packet loss because of rate limiting policy set previously. Running Cisco 3650. I've removed the policies from each of the port's VLANs assigned over an hour ago and Observium is still showing as the policies in place? It polled ample times on all ports but not quite sure why it's still showing the policies and errors? I did notice at least that it showed no packet loss once I removed the policies, but the error bubble is still there and it still shows the policy name when viewing the CBQoS page for the ports. >From the Switch Switch2#conf t Enter configuration commands, one per line. End with CNTL/Z. Switch2(config)#int gi2/0/14 Switch2(config-if)#no service-policy output speed25 % policy map speed25 not configured Switch2#sh platform qos policies port Policy: None in the list Switch2#show run interface gi2/0/14 Building configuration... Current configuration : 113 bytes ! interface GigabitEthernet2/0/14 description Server 19 switchport access vlan 19 switchport mode access end

1 0

alerts
by Célia LAMY 04 Oct '16

04 Oct '16

Hello, I created three alerts for all of my servers. I'm not getting the alerts both servers. For others it works. How to do ? I use Observium CE 0.16.1.7533. Cordialement -- Célia LAMY Service Informatique ACTES SUD Place Nina Berberova, 13200 Arles Tél : 04 88 65 90 25

5 6

Billing update
by Eduardo Schoedler 04 Oct '16

04 Oct '16

Hi, I just added a port to a billing, so I imagine observium is going to calculate since today, right? However, today is day 3. How can I generate statistics for a specific billing since day 1? Thank you. Regards, -- Eduardo Schoedler

3 7

Alert on failed HDD
by Fredrik Winäs 04 Oct '16

04 Oct '16

Hi Did a test on a Linux (Ubuntu 16 on a Supermicro SYS-6017R-73HDP+) machine to see if we could get an alert when a HDD fails. The machine isn't using any RAID , just a plain mount of a partition to a mount point. I have configured IPMI in Observium to. The test (a bit brutal..) was to just remove the HDD when the server was ON. I did not see anything in Observium about the missing drive. What can I do so I can set up an alert checker notifying me about a failed drive? Will the unix_agent give me more info on this? Med Vänlig Hälsning Fredrik Winäs tel +46 31 762 02 43 fredrik.winas(a)cgit.se<mailto:fredrik.winas@cgit.se> www.cgit.se<http://www.cgit.se/> CGit AB Taljegårdsgatan 11C SE-431 53 Mölndal [cid:image001.png@01D21E53.0889E7A0]

1 0