On Tue, Apr 13, 2010 at 02:06:37AM -0400, [email protected] wrote:
>   Open vswitch seems to be working on a logic of matching flow hash
>   with that of IP data packets.  But the role of flow hash bucket in
>   user space and kernel space is an area to explore in.  Can anyone
>   put more light on the role of flow hash matching used in open
>   vswitch, the parameters used in flow hash calculation ? are there
>   two separate flow hash tables maintained at kernel space and user
>   space ?

The kernel extracts flow data (struct odp_flow_key) from each received
packet and hashes the data using a Jenkins hash.

User space maintains a more general flow classifier that supports
wildcards.  It also uses a Jenkins hash as part of the classifier.

_______________________________________________
discuss mailing list
[email protected]
http://openvswitch.org/mailman/listinfo/discuss_openvswitch.org

Reply via email to