Forum Discussion

Simon_Waters_13's avatar
Simon_Waters_13
Icon for Cirrostratus rankCirrostratus
Nov 27, 2014

APM Session log - "no leasepool assigned" - intermittent

We have a problem with HA pair, there is a support ticket open.

 

It is intermittent, when it occurs the Big-IP Edge Client disconnects with the error "no leasepool assigned" at the end of the session report.

 

Attempts to create an archive (on the current active device, with keys and encryption) cause the web interface to hang. It looks like the archive is created, but due to the web interface issues this is not apparent.

 

Simply reapplying the relevant Access Policy (one character change) causes everything to return to normal.

 

Hints of what to look for appreciated.

 

5 Replies

  • Hello,

     

    It seems that you reach the limit of your lease pool.

     

    Can you check how many IPs are available in the lease pool and how many concurrent sessions you may have ?

     

    Yann

     

  • Hi Yann,

     

    the IPv4 lease pool is a randomly selected /24 in 10.x.x.x, fourth octet 1 to 253, so I should have 253 IP addresses to play with.

     

    The network access in APM Dashboard suggests I've had at most one concurrent connection in the last week. This is used for management access only, so I'd be very worried if it got to double figures for concurrent connections.

     

    End user access is via Access Policies and regular virtual servers with no leasepools involved.

     

    Is there a specific measure of leasepool usage? e.g. Can it leak? How would I tell?

     

    I wouldn't expect this to impact on the ability to make encrypted archives if it was simple exhaustion, but I guess reapplying a policy might reset more than one problem.

     

  • marco's avatar
    marco
    Icon for Nimbostratus rankNimbostratus

    Hello Simon,

     

    Which version are you using ? I might have the same problem on a cluster in 11.6 HF1. Depending on the cluster member, it seems that the lease pool is not really taken into consideration.

     

    Marco

     

  • 11.6.0 HF1 but I saw this issue before I got there.

     

    I think something basic is wrong with the configuration, as it is not working at all on the other one of the HA pair currently.

     

  • what do you see in your APM logs when it happens? Such entries?

     

    IPv4 Addr collision: