On Tue, Jun 18, 2024 at 3:41 PM Dumitru Ceara <dce...@redhat.com> wrote: > > It improves the debugging experience if we can easily get a list of > OpenFlow rules and groups that contribute to the creation of a datapath > flow. > > The suggested workflow is: > a. dump datapath flows (along with UUIDs), this also prints the core IDs > (PMD IDs) when applicable. > > $ ovs-appctl dpctl/dump-flows -m > flow-dump from pmd on cpu core: 7 > ufid:7460db8f..., recirc_id(0), .... > > b. dump related OpenFlow rules and groups: > $ ovs-appctl upcall/dump-ufid-rules ufid:7460db8f... pmd=7 > cookie=0x12345678, table=0 > priority=100,ip,in_port=2,nw_dst=10.0.0.2,actions=resubmit(,1) > cookie=0x0, table=1 priority=200,actions=group:1 > group_id=1,bucket=bucket_id:0,actions=ct(commit,table=2,nat(dst=20.0.0.2)) > cookie=0x0, table=2 actions=output:1 > > The new command only shows rules and groups attached to ukeys that are > in states UKEY_VISIBLE or UKEY_OPERATIONAL. That should be fine as all > other ukeys should not be relevant for the use case presented above. > > For ukeys that don't have an xcache populated yet, the command goes > ahead and populates one. In theory this is creates a slight overhead as > those ukeys might not need an xcache until they see traffic (and get > revalidated) but in practice the overhead should be minimal. > > This commit tries to mimic the output format of the ovs-ofctl > dump-flows/dump-groups commands. For groups it actually uses > ofputil_group/_bucket functions for formatting. For rules it uses > flow_stats_ds() (the original function was exported and renamed to > ofproto_rule_stats_ds()). > > Signed-off-by: Dumitru Ceara <dce...@redhat.com> > ---
Acked-by: Mike Pattrick <m...@redhat.com> _______________________________________________ dev mailing list d...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-dev
