Forum Discussion

Scott_C_16492's avatar
Scott_C_16492
Icon for Nimbostratus rankNimbostratus
Feb 25, 2009

Failing Monitors

I'm having a problem with the builtin monitors on a pair of LTMs

 

 

Pool members keep being marked as down by the builtin HTTP/HTTPS monitor:

 

 

Feb 25 14:16:07 dhlb02 mcpd[1811]: 01070638:5: Pool member 192.168.147.36:7080 monitor status down.

 

Feb 25 14:16:07 tmm tmm[1626]: 01010028:3: No members available for pool wiki_wf_intranet_http

 

 

 

But when I telnet from the LTM I get a connection:

 

 

[root@dhlb02:Active] run telnet 192.168.147.36 7080

 

Trying 192.168.147.36...

 

Connected to 192.168.147.36.

 

Escape character is '^]'.

 

GET / HTTP/1.0

 

 

HTTP/1.1 301 Moved Permanently

 

Date: Wed, 25 Feb 2009 14:17:23 GMT

 

Server: Apache/2.0.55 (Unix) PHP/5.2.0 mod_ssl/2.0.55 OpenSSL/0.9.7e-p1

 

Last-modified: Wed, 25 Feb 2009 14:17:24 GMT

 

 

 

Couple of minutes later, the monitor marks it as back up.

 

 

Any ideas?
management
monitoring

1 Reply

Sort By
  • hooleylist's avatar
    hooleylist
    Icon for Cirrostratus rankCirrostratus
    Feb 25, 2009
    Are you using a receive string in either monitor? If not, you're basically performing a check to see if the server completes a three way TCP handshake. I'd suggest using an HTTP 1.1 formatted request for both the HTTP and HTTPS monitors and configure a receive string to ensure the pool member is responding. Here are examples for the send/receive strings:

     

     

    send:

     

    GET /index.jsp HTTP/1.1\r\nHost: \r\nConnection: Close\r\n

     

     

    receive:

     

    200 OK

     

     

    If you still see the monitor marking down pool members you can use tcpdump to check the communication between LTM and the pool members:

     

     

    tcpdump -ni 0.0 -s0 -w/var/tmp/monitor.dmp "host LTM_STATIC_SELF_IP and (host SERVER1 or host SERVER2)"

     

     

    Replace LTM_STATIC_SELF_IP with the non-floating (static) self IP address on the server VLAN. Replace SERVER1 and SERVER2 with the pool member IP addresses.

     

     

    You can use WinSCP to copy the tcpdump file from LTM to a workstation and then Wireshark to analyze the tcpdump. You can check SOL411 (Click here) and SOL1893 (Click here) for details on analyzing a tcpdump.

     

     

    You can also enable debug on the monitoring daemon, bigd, by running 'b db bigd.debug enable' from the command line. The output is logged to /var/log/bigdlog. Make sure to disable debug when you're done by running 'b db bigd.debug enable'.

     

     

    Aaron