Forum Discussion

jan_de_wachter_'s avatar
jan_de_wachter_
Icon for Nimbostratus rankNimbostratus
Jul 27, 2015

LTM: Policy problem

Does anybody has the same problem or has someone a solution.

 

We have a virtual server active with iRule and Policies activated.

 

For a lot of (automatic generated by iApp) applications which DNS point to this virtual server we created policy rules to change the URI when needed and to forward to an application specific virtual server.

 

This application specific virtual server has specific profiles an a specific pool.

 

This is working perfect. BUT sometimes we get calls form the customer that the application is not working. They get a message from the DNS referenced virtual server saying that there is no server active for a blanco pool! (generated by the DNS selected virtual server iRule). What we at that moment is the following: - Virtual server active - Specific virtual server is active - Application pool is active - In the statistics we see that the uri replacenment and the virtual server forwarding has been selected. - All applications making use of the policy rules are having the same problem.

 

When we do a REORDER of the policy rules - making the update button available - and pushing UPDATE (without making any modifications) all is again running normally (I presume forcing a reload of the policies).

 

Has anyone had the same problem and / or how can we solve this problem I already opened a case for support but did not yet get any answer.

 

Thanks

 

application delivery
deployment
LTM

1 Reply

Sort By
  • jan_de_wachter_'s avatar
    jan_de_wachter_
    Icon for Nimbostratus rankNimbostratus
    Jul 31, 2015

    This particular problem has been solved. It was an internal problem. The virtual server default pool was deleted. Our forwarding policy did a forward to virtual server (pool was set to spaces). In our irule HTTP_REQUEST we felt in the default option (url was not defined in iRule - only in policy), where we did a test comparing the pool with the default pool, which was positive and resulted in the error message.

     

    Seen this problem we restored the situation (and also the default pool setting) - all OK.

     

    We also did a reorder-update of the policies because we had before problems when the "forwarded to" virtual server was not active, which resulted in a LB_FAILED condition. When the virtual seerver became again active, we still got the LB_FAILED condition. To resolve this problem we had to do a reorder-update of the policy rules.