For more information regarding the security incident at F5, the actions we are taking to address it, and our ongoing efforts to protect our customers, click here.

Forum Discussion

Heiko_Engemann_'s avatar
Heiko_Engemann_
Icon for Nimbostratus rankNimbostratus
Jun 13, 2008

cacti stopped graphing

Hello,

i followed the guide on this site to setup a cacti monitoring,,

it worked great for some weeks, but than it stopped wokring.

All the graphs are empty for 2 weeks now.

We have 2 BigIP Servers, the active one seems not to response to any snmp requests.

We created a new VS and on the passive I see in the list where I can add new Graphs,

but not on the active one.

here is a snmp example: 

       
        [root@monitoring-1:/var/www/cacti/log] snmpwalk -v 2c -c public 172.17.1.12 .1.3.6.1.4.1.3375.2.2.10.2.3.1.12.18.119.119.119.46.103.115.45.119.97.114.114.97.110.116.115.46.100.101        
        SNMPv2-SMI::enterprises.3375.2.2.10.2.3.1.12.18.119.119.119.46.103.115.45.119.97.114.114.97.110.116.115.46.100.101 = No Such Object available on this agent at this OID     
   

        
        [root@monitoring-1:/var/www/cacti/log] snmpwalk -v 2c -c public 172.17.1.11 .1.3.6.1.4.1.3375.2.2.10.2.3.1.12.18.119.119.119.46.103.115.45.119.97.114.114.97.110.116.115.46.100.101        
        SNMPv2-SMI::enterprises.3375.2.2.10.2.3.1.12.18.119.119.119.46.103.115.45.119.97.114.114.97.110.116.115.46.100.101 = Counter64: 0

a simple snmpwalk withtout an OID brings a a couple of results.

this are the errors in the cacti logfile: 

              
        06/13/2008 08:15:21 AM - CMDPHP: Poller[0] Host[9] DS[434] WARNING: Result from SNMP not valid.  Partial Result: No Such Object avail        
        06/13/2008 08:15:21 AM - CMDPHP: Poller[0] Host[9] DS[435] WARNING: Result from SNMP not valid.  Partial Result: No Such Object avail        
        06/13/2008 08:15:21 AM - CMDPHP: Poller[0] Host[9] DS[439] WARNING: Result from SNMP not valid.  Partial Result: No Such Object avail        
        06/13/2008 08:15:21 AM - CMDPHP: Poller[0] Host[9] DS[434] WARNING: Result from SNMP not valid.  Partial Result: No Such Object avail        
        06/13/2008 08:15:21 AM - CMDPHP: Poller[0] Host[9] DS[436] WARNING: Result from SNMP not valid.  Partial Result: No Such Object avail        
        06/13/2008 08:15:21 AM - CMDPHP: Poller[0] Host[9] DS[435] WARNING: Result from SNMP not valid.  Partial Result: No Such Object avail

I created 2 news devices, one for each LB, in associated data queries, there is a difference between both LB: 

        
 1) F5 BigIP - Interface Statistics   (Verbose Query)   Uptime Goes Backwards   Success 14 Items, 7 Rows    
 2) F5 BigIP - VS Statistics (Verbose Query) Uptime Goes Backwards Success 452 Items, 113 Rows    
 

                 
 1) F5 BigIP - Interface Statistics   (Verbose Query)   Uptime Goes Backwards   Success 14 Items, 7 Rows  
 2) F5 BigIP - VS Statistics (Verbose Query) Uptime Goes Backwards Success 0 Items, 0 Rows

Strange thing is that it also does not talk to nagios, the passive one does.

I cant find any difference in the settings on both servers, we didnt change anything.

Can this be a load problem?

thx

Heiko
management
monitoring

6 Replies

  • JRahm's avatar
    JRahm
    Icon for Admin rankAdmin
    Jun 16, 2008
    You're not the first to experience this issue. Open a case with support, as both units should always answer queries. You could try to kickstart it by restarting the snmp daemon (bigstart restart snmpd)
  • Heiko_Engemann_'s avatar
    Heiko_Engemann_
    Icon for Nimbostratus rankNimbostratus
    Jun 17, 2008
    hello elah,

     

    a restart of the snmpd did help, a bit.

     

    Now the snmpwalk does work on both, but cacti still is only creating some graphs.

     

    Il wait an hour and see what happens.

     

    EDIT:

     

     

    It is working with some breaks, it stops for some minutes and than starts graphing again.

     

    With our nagios I get all the Vserver stats but no poolstats for our active unit.
  • Hamish's avatar
    Hamish
    Icon for Cirrocumulus rankCirrocumulus
    Jun 18, 2008
    Ahh.... What version of BigIP you running? We used to have problems back with one of the 4.x branches where snmpd had a memory leak... It would go slower & slower, eventually only responding after the SNMP manager had timed out (30 seconds in that case IIRC).

     

     

    I know haven't seen it at 9.3.1, not sure if we saw it at 9.1 or not...

     

     

    H
  • Heiko_Engemann_'s avatar
    Heiko_Engemann_
    Icon for Nimbostratus rankNimbostratus
    Jun 23, 2008
    Hello,

     

     

    it is now working for the last days. thx for the ip.

     

    We still have problems with snmp queries and nagios.

     

    I have a plugin which monitors the poolmembers, when there are not enoguh we get a warning.

     

    Are there any other linux commadline tools that can monitor the state of a bigIP pool and report

     

    changes?

     

     

    thx

     

     

    HEiko
  • JRahm's avatar
    JRahm
    Icon for Admin rankAdmin
    Jun 23, 2008
    Check out the codeshare, there are plenty of examples (or close) in a couple languages that will get you on the way. Here's an example of pool status in pyControl:

     

     

    http://devcentral.f5.com/wiki/default.aspx/iControl/pyControlQuickPoolStatus.html Click here
  • Heiko_Engemann_'s avatar
    Heiko_Engemann_
    Icon for Nimbostratus rankNimbostratus
    Jun 25, 2008
    Hello,

     

    well im not familiar with python, and in the example are some lines missing i can try that out.

     

    Are there any out of the box monitoring tools for F5 LoadBalancers,

     

    we need something that give us an email when a poolmembers isnt available any more.

     

     

    Other questions is, can I give a snmp query on the bigIP more priorioty?

     

    I still get timeouts on the pool stats, tried to run the nagios plugin from the bash, but still there are some timeouts.

     

     

     

    greetings