- observium - lists.observium.org

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 05 Oct '16

05 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 04 Oct '16

04 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 04 Oct '16

04 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

Custom OID in observium community edition
by Janusz Janik 03 Oct '16

03 Oct '16

Hi, Is there any way to enable _custom oid_ option in observium community edition menu? Or is this option requires subscription version of observium? PS We use observium: > > Version Information > > *Observium CE* 0.16.9.8118 (30th September 2016) > Thank you for your response. Best regards, Janusz Janik -- Pozdrawiam, Janusz Janik

2 1

SLA : Search box/global search?
by Johann Mallet 30 Sep '16

30 Sep '16

Hello, It's possible to activate the SLAs search from the global search box? Another question : Search box in the SLA page is planned in the future? We have 2040 SLA (and many more in the future (>10k)). It's complicated for the support's guys :-) Thanks, Johann

2 2

Custom alerts without using checkers
by Andy Lemin 30 Sep '16

30 Sep '16

Hi, We have many custom scripts, and we would like those scripts to be able to send custom alerts into Observium. What is the best way to do this? We would not be able to use "Alert checks".. Or is there a special trap that we can use for this where we can define the Alert message string in the trap packet? Cheers, Andy.

1 0

Responsive Mode
by chad＠talkjesus.com 30 Sep '16

30 Sep '16

Anyone familiar with any possible mobile friendly, responsive them or even app? I searched intently and found only one app that hasn't been updated in years.

2 1

Wowza MIB
by Matt Vinson 29 Sep '16

29 Sep '16

How does one go about requesting MIB support? Or alternatively, adding MIB support manually to one's own server? Wowza Streaming Engine just recently added SNMP and developed their own MIB. It would be extremely useful for us to be able to monitor this application in depth. Documentation: https://www.wowza.com/forums/content.php?831-How-to-use-SNMP-in-Wowza-Strea… MIB: https://www.wowza.com/downloads/forums/snmp/WowzaStreamingEngineMIB.zip -- Regards, Matt Vinson TSS Digital Services matt.vinson(a)tssdigital.com 360.915.1100 --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus

4 4

email alert template tuning
by Boyard Engels 29 Sep '16

29 Sep '16

Hey Team, I am keen to reduce the information that is sent from Observium via the email alerts. Is there any information available on how to tune the alert emails? Thanks!! Bo

1 0

High CPU load on Cisco S500G Small business switches
by Jamie Tynan 29 Sep '16

29 Sep '16

Hi, We've noticed an issue with Cisco SG500 switches and Observium, where if you enable stacking mode (the default on the switch), regardless of if it is in a stack or not, the CPU reports a 50% average. Our investigations show it runs at 95% + for a couple of minutes whilst Observium is polling it, then drops back down to <10% when the query finishes. Turning stacking mode off drops the average CPU to <10% and CPU does not ramp up and down. We raised a case with Cisco TAC, who said this is "expected behaviour": >>> At every 5 min cycle, the CPU rises to 85- 95% temporarily and then drops down. The reason for this is, Observium is checking for interface status. When switch is in standalone mode the interface count database is the max switch capacity. Example 48 ports = 48 ports plus channel ports ~ 60 ports In stack mode the predefined port count in the data base is around 880 ports. While the switch is responding to SNMP get , it is using up resources and that is why the CPU utilization will be high. While I was testing it, I did not see or encountered any network outage or Web access issues. We have already made an enhancement request for this behaviour, to limit the port count to any stack configuration. >>> Has anyone else encountered anything similar? Adam - I realise that Observium is only polling what the switch is reporting, but if Observium can ascertain how many switches are in the stack, would it be possible to limit polling to only query active/connected stack members? Thanks, Jamie

2 4

Alcatel Lucent : Strange "Interface changed: [ifConnectorPresent]" flood
by Johann Mallet 29 Sep '16

29 Sep '16

Hello, I have many Alcatel Lucent routers in my backbone. Since I have add these routers, my eventlog is flooded by "Interface changed: [ifConnectorPresent]": http://img15.hostingpics.net/pics/712223floodalcatel.png I have 1274 eventlog pages for the router in example :-/ Do you have the same problem? Thanks, Johann

3 6

Using Obserivum on server with multiple interfaces/vlan?
by Johann Mallet 29 Sep '16

29 Sep '16

Hello, Simple question : We need upgrade our Observium deployment with multiple networks (public & private). Does Observium have any issue accessing nodes via more than one server interface? Whether two different physical ports or two different vlans on the same port? Some proprietary vendor NMS solutions are somehow locked to one interface only and cause great trouble. :o) Thanks, Johann

2 2

Bandwidth Reporting Off
by Talk Jesus 28 Sep '16

28 Sep '16

I’ve been using Observium for a while and there seems to be some discrepancies between my graphs and the upstream (my data center provides me Ethernet GigE bandwidth). For example, from around 8am to 12pm today it showed overall utilization pretty consistent well above 400Mbps. These switches are actually trunked (2 x 3650 L3 switches). In the graph, it shows the uplink to switch #2 connectivity calculated with the total utilization (VLAN2155). This is definitely not accurate as it shouldn’t do that as far I’m concerned. Total utilization should simply be Drops #1 & #2, which covers all VLANs on both switches. Example: Drop #1 = 40Mbps Drop #2 = 200Mbps Uplink to Switch #2 = 130Mbps Now as a result VLAN2155 (total utilization) is calculating not only #1 and #2 drops from the upstream, but even the trunk port labeled “Uplink to Switch #2”. I disabled polling on that uplink port and the total utilization (VLAN2155) is still far higher than what the output is for the past 24 hours from the upstream provider. They show 187Mbps past 24 hours (95th), 219M In / 202 Out. I show 266M In / 182M Out. Why is it so drastically different? Why is Observium calculating the “uplink” to the trunked 2nd switch in the total utilization graph? This is throwing off accuracy.

5 9

VRF/BGP/OSPF discovery in vrf-lite (nx-os)
by Stef Renders 28 Sep '16

28 Sep '16

Hi Does anyone know whether it is currently possible to discover bgp/ospf routing instances within a VRF on the NX-OS platform ? We migrated from IOS to NX-OS, IOS was passing them through snmp just fine. But NX-OS requires snmpv3 contexts for vrf resources. A full walk in the 'default-vrf' does not contain any metrics for the vrf's running on this VDC. The problem is more or less described here as well: https://github.com/librenms/librenms/issues/280 Kind regards Stef

4 6

Port Group / Entity Question
by Pico Leto 28 Sep '16

28 Sep '16

Hi, Can a port group contain multiple entities checks? I'm trying to just group our carriers for each location into port group. For instance I have our isp/carriers listed in the following format for the port descriptions on our routers L3:CKTID:DEMARC COG:CKTID:DEMARC My entity matches on either of those but when I combine them like the following below, I don't see any devices being associated with it. ifAlias matches *COG:* ifAlias matches *L3:*

2 1

Update Error
by Talk Jesus 28 Sep '16

28 Sep '16

Trying to update my CE from 15x to the latest http://www.observium.org/docs/updating/ Ran: cd /opt mv observium observium_old wget -Oobservium-community-latest.tar.gz http://www.observium.org/observium-community-latest.tar.gz tar zxvf observium-community-latest.tar.gz mv /opt/observium_old/rrd observium/ mv /opt/observium_old/*log* observium/ mv /opt/observium_old/config.php observium/ [root@rtg opt]# /opt/observium/discovery.php -u PHP Fatal error: require(): Failed opening required '/opt/observium/config.php' (include_path='.:/usr/share/pear:/usr/share/php') in /opt/observium/includes/sql-config.inc.php on line 26

4 9

Can't Hide Search
by Talk Jesus 28 Sep '16

28 Sep '16

Upgraded to latest CE version yesterday. Now the layout is different from 15x and the big search box on the left side of the port listing or device listing won't hide. I replicated this 2 browsers as well.

3 3

Re: [Observium] Cisco subinterfaces
by ron＠rjr-services.com 28 Sep '16

28 Sep '16

Yeah, that command. ;-) Get Outlook for Android On Wed, Sep 28, 2016 at 7:46 AM -0500, "Markus Klock" <markus(a)best-practice.se> wrote: -- This message has been scanned by E.F.A. Project and is believed to be clean.

1 0

Cisco subinterfaces
by Mirditor Kuçaj 28 Sep '16

28 Sep '16

Hello Lost the graphs on subinterfaces and portchannels. We rebooted a 7604 with IOS 15.4(3)S2 and monitored with Observium 0.16.9.8104. Somehow the graphs ID-are changed and doesnt belong to the same subinterface/portchannel. Thx Mirditor Kuçaj Backbone miri(a)abissnet.al <mailto:miri@abissnet.al> Abissnet Sh.a. +355 69 60 91 364 +355 (0)44 300 003 Ext 139 Rr. Ismail Qemali, Pallati Abissnet, Tiranë <http://www.abissnet.al/> www.abissnet.al <https://www.facebook.com/Abissnet> <https://plus.google.com/u/0/+AbissnetalISP> <https://www.linkedin.com/profile/view?id=AAIAABqLWRAB88ljU6t8b67oVm8PAlvPrh mPybE&trk=nav_responsive_tab_profile%20>

2 1

Re: [Observium] Cisco subinterfaces
by ron＠rjr-services.com 28 Sep '16

28 Sep '16

Please review this Cisco document regarding interface persistence. This is likely the issue you are seeing, not a problem with Observium itself. http://www.cisco.com/c/en/us/support/docs/ip/simple-network-management-prot… There is a command you can include in your configuration that will keep interface numbering persistent across reboots. I just don't recall it exactly off the top of my head, or I would include it here. ... Ron Get Outlook for Android On Wed, Sep 28, 2016 at 7:34 AM -0500, "Mirditor Kuçaj" <miri(a)abissnet.al> wrote: -- This message has been scanned by E.F.A. Project and is believed to be clean.

1 0

Trunk Issues
by Talk Jesus 28 Sep '16

28 Sep '16

I've been using Observium for a while and there seems to be some discrepancies between my graphs and the upstream (my data center provides me Ethernet GigE bandwidth). For example, from around 8am to 12pm today it showed overall utilization pretty consistent well above 400Mbps. I have two drops from them: primary and secondary (redundant). Plus, they both show active instead of active/passive. Not sure if that has anything to do with it. >From the tech: "95th percentile is calculated from both ingress and egress traffic. For example, if you used a constant 200Mbps up and down, then you would only be charged your commit of 200Mbps, not double. I do see what you mean that the two handoffs (layer 2) are equaling more than 300Mbps, but when you look at the Aggregate (layer 3) graph at the top, it does not go over 300Mbps. These packets can possibly be queued before going out so you don’t go over your commit. " So, what do you think might be causing this hiccup? I noticed this too: so these switches are actually trunked (2 x 3650 L3 switches). In the graph, it also shows the uplink connectivity calculated with the total utilization. This is definitely not accurate as it shouldn't do that. Example: Drop #1 = 40Mbps Drop #2 = 200Mbps Uplink to Switch #2 = 130Mbps Now as a result VLAN2155 (total utilization) is calculating not only #1 and #2 drops from the upstream, but even the trunk port labeled "Uplink to Switch #2". How can I rectify this with Observium to avoid that and get an accurate output?

5 11