Forum Discussion

OhHwanYoung_297's avatar
OhHwanYoung_297
Icon for Nimbostratus rankNimbostratus
Aug 10, 2017

How Carp algorithm with source address persistence works?

what is the difference of source address persistence only and Carp algorithm with source address persistence?

 

1 Reply

  • The default source-address persistence (no CARP hashing) creates a record on the BIG-IP that associates a source address (or range, based on the mask configured) with a pool member. You can view the current source-address persistence records in TMSH:

    show ltm persistence persist-records mode source-address
    

    The CARP hashing method is based on the Cache Array Routing Protocol. In this method, the source address is an input into a hashing function. The result of the hashing function is used to select the pool member that the traffic will be sent to. The hashing function should compute the same result every time given the same input (source address), so traffic from the same source/mask will always be sent to the same pool member. If a pool member goes down, the system can reallocate hash results across the remaining pool members.

    Given that the CARP hashing method is computed and not records-based, you cannot display the persistence results. I'm not aware of a way to display the hash allocations to pool members on LTM, and last I checked F5's specific implementation of CARP hashing is not publicly available.

    Edited to add: Some good additional information in the discussion of this post.