On 2/11/22 11:49, Stokes, Ian wrote: >>> -----Original Message----- >>> From: dev <[email protected]> On Behalf Of Kumar Amber >>> Sent: Wednesday 9 February 2022 09:50 >>> To: [email protected] >>> Cc: Amber, Kumar <[email protected]>; [email protected] >>> Subject: [ovs-dev] [PATCH v3] system-dpdk.at: Add warning log in mfex fuzzy >> test. >>> >>> Some specific warning are seen on various systems >>> which may not be visible on others but good to add >>> such logs to test to avoid test-case failure. >>> >>> Thw warning only effects the fuzzy tests due to >>> more than 1000+ flows being offloading simultanously. >>> >>> Wilcarding flow count number as for different systems >>> under test the number could vary in the warning log. >>> >>> Suggested-by: Eelco Chaudron <[email protected]> >>> Signed-off-by: Kumar Amber <[email protected]> >>> >> >> I didn't see the issue that this patch fixes initially. However, if I make >> the below >> change, I can see the error during the test. >> >> Change: >> diff --git a/tests/system-dpdk.at b/tests/system-dpdk.at >> index 9384cf7f0..165b68f0e 100644 >> --- a/tests/system-dpdk.at >> +++ b/tests/system-dpdk.at >> @@ -256,6 +256,8 @@ AT_KEYWORDS([dpdk]) >> AT_SKIP_IF([! $PYTHON3 -c "import scapy"], [], []) >> AT_CHECK([$PYTHON3 $srcdir/mfex_fuzzy.py $srcdir], [], [stdout]) >> OVS_DPDK_START() >> +AT_CHECK([ovs-vsctl --no-wait set Open_vSwitch . other_config:flow- >> limit=13500]) >> + >> >> dnl Add userspace bridge and attach it to OVS >> AT_CHECK([ovs-vsctl add-br br0 -- set bridge br0 datapath_type=netdev]) >> >> Error: >>> 2022-02-10T09:42:11.350Z|00006|ofproto_dpif_upcall(pmd- >> c21/id:101)|WARN|upcall: datapath reached the dynamic limit of 13500 flows. >> >> Which causes a fail: >> 7. system-dpdk.at:254: FAILED (system-dpdk.at:280) >> >> Applying this patch while keeping the above diff to generate the flow limit >> error >> results in the test passing. I can see the "reached the dynamic limit" >> warning in >> the ovs-vswitchd.log but it doesn't cause a test failure. >> >> Finding the value of "13500" as a flow-limit took some experimentation and >> would probably vary based on platform. >> >> With that, and looking at the .at changes, LGTM. >> >> Acked-by: Cian Ferriter <[email protected]> > > Thanks for testing and reviewing folks. > > So does it make sense to apply this to both 2.17 and 2.16 also?
Since it's a test-only change, and a fix actually, it should be OK to backport. Best regards, Ilya Maximets. _______________________________________________ dev mailing list [email protected] https://mail.openvswitch.org/mailman/listinfo/ovs-dev
