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

Muhammad_Irfan1's avatar
Muhammad_Irfan1
Icon for Cirrus rankCirrus
Feb 26, 2015

Client server is getting exceptions when trying to access F5 VIP.

We have F5 LTM 5000s 11.4.1 in production. I have a VS and proxypass iRule is applied to it which means one VIP is working for many pools (services).

Everything was running fine for 2 months. Among many services only one service is not having some problem. The problem is that there is one server (acting as client for F5) which access this VIP, get load balance between 3 servers below F5, The server(client) is now getting exceptions 3,4 times a day while accessing VIP. 

 2015-02-26 12:13:34,928 INFO
 [com.securetech.bioserver.operator.service.impl.OperatorService] -  
  <$59951987$..PreSale...Exeception..Could not invoke service.. Nested exception is 
      org.codehaus.xfire.fault.XFireFault: Couldn't send message....NewSales..$CNIC$..3120516529777..$Id$..59951987...$msisdn$..03006355991...
      $imsi$..00970...$imei$..10013747...UserMsisdn..0300868
     org.codehaus.xfire.XFireRuntimeException: Could not invoke service.. 
     Nested exception is org.codehaus.xfire.fault.XFireFault: Couldn't send message.

Remember traffic from client server till F5 is https and from F5 to pool server is http.

What could be the reason of these exceptions?

3 Replies

  • StephanManthey's avatar
    StephanManthey
    Icon for Nacreous rankNacreous
    Mar 01, 2015
    Hi Muhammad, are you observing simultaneous log messages in the /var/log/ltm facility? Recent logs can be viewed via "zcat". Due to the log rotation logs of last 10 days are kept, as you probably know. As it is an error on the client it´s hard to tell, what it actually means. Perhaps the developer of the client software can explain potential reasons. The "couldn´t send message" log indicates the request is not forwarded to the target server. If the /var/log/ltm reports a simultaneous pool member state change or a "http header size exceeded" we would have an explanation. Duplicated headers may be another reason depending on the iRule logic. Thanks, Stephan
  • Muhammad_Irfan1's avatar
    Muhammad_Irfan1
    Icon for Cirrus rankCirrus
    Mar 01, 2015
    HI Stephan, This is the correlated issue of pool members going down while application running. Client server access those pool members. This service takes about 30 thousands requests per second about 4 thousands per 10 minutes. Last night when pool members were going up down up down, I asked application guy how much requests were you getting during that time and he said there are gaps of 2 minutes in between the requests. Definitly when F5 will mark pool members down it will not get requests and client server will see exceptions. Now we have to find the reason why pool members are marked down. Does application hangs or due to traffic load it is unable to respond to monitor requests.