dpif-netdev: Change pmd selection order.

Up to his point rxqs are sorted by processing cycles they
consumed and assigned to pmds in a round robin manner.

Ian pointed out that on wrap around the most loaded pmd will be
the next one to be assigned an additional rxq and that it would be
better to reverse the pmd order when wraparound occurs.

In other words, change from assigning by rr to assigning in a forward
and reverse cycle through pmds.

Also, now that the algorithm has finalized, document an example.

Suggested-by: Ian Stokes <ian.stokes@intel.com>
Signed-off-by: Kevin Traynor <ktraynor@redhat.com>
Signed-off-by: Darrell Ball <dlu998@gmail.com>

1 files changed, 16 insertions, 0 deletions

diff --git a/Documentation/howto/dpdk.rst b/Documentation/howto/dpdk.rst
index a67f3a1fa..bac51de8e 100644
--- a/Documentation/howto/dpdk.rst
+++ b/Documentation/howto/dpdk.rst
@@ -123,6 +123,22 @@ automatically. The processing cycles that have been stored for each rxq
 will be used where known to assign rxqs to pmd based on a round robin of the
 sorted rxqs.
 
+For example, in the case where here there are 5 rxqs and 3 cores (e.g. 3,7,8)
+available, and the measured usage of core cycles per rxq over the last
+interval is seen to be:
+
+- Queue #0: 30%
+- Queue #1: 80%
+- Queue #3: 60%
+- Queue #4: 70%
+- Queue #5: 10%
+
+The rxqs will be assigned to cores 3,7,8 in the following order:
+
+Core 3: Q1 (80%) |
+Core 7: Q4 (70%) | Q5 (10%)
+core 8: Q3 (60%) | Q0 (30%)
+
 Rxq to pmds assignment takes place whenever there are configuration changes.
 
 QoS