Hi Ben, They belong to ICMP flows from 10.0.0.2 to 10.0.0.3 and vice-versa. The timeout disappeared as for R2, I have specified only hard_timeout. I assumed that since hard timeout duration elapsed so timeout disappeared. But this is weird, conceptually R1 should be there instead of R2 as it has no hard timeout, right? I am using OVS switch with mininet and POX controller in Ubuntu system.
On Tue, Jan 10, 2017 at 1:55 PM, Ben Pfaff <b...@ovn.org> wrote: > On Tue, Jan 10, 2017 at 01:49:51PM -0500, Quamar Niyaz wrote: > > Hi All, > > I have following scenarios for my network application. I install a rule > R1 > > with a flow id F1, action output to port 2 and 4, idle timeout 60s, and > > priority 1000. After few seconds, I install another rule R2 with a flow > id > > F1, action output to port 2 only, hard timeout less than 60s, and > priority > > 2000. > > > > What I was assuming that R2 should expire due to hard timeout and R1 be > > there in the flow table. However, I found R1 expires and R2 persists in > the > > table. Can anyone explain it? > > > > *After R1 and R2 installation:* > > n_packets=9,n_bytes=882,hard_timeout=47,priority=2000,icmp, > nw_src=10.0.0.3,nw_dst=10.0.0.2,icmp_type=0,icmp_code=0 > > actions=output:3 > > n_packets=10,n_bytes=980,idle_timeout=60,priority=1000,icmp, > nw_src=10.0.0.3,nw_dst=10.0.0.2,icmp_type=0,icmp_code=0 > > actions=output:3,4 > > n_packets=9,n_bytes=882,hard_timeout=47,priority=2000,icmp, > nw_src=10.0.0.2,nw_dst=10.0.0.3,icmp_type=8,icmp_code=0 > > actions=output:2 > > n_packets=10,n_bytes=980,idle_timeout=60,priority=1000,icmp, > nw_src=10.0.0.2,nw_dst=10.0.0.3,icmp_type=8,icmp_code=0 > > actions=output:2,4 > > > > *After 60s R1 expires and R2 persists:* > > n_packets=255,n_bytes=24990,priority=2000,icmp,nw_src=10. > 0.0.3,nw_dst=10.0.0.2,icmp_type=0,icmp_code=0 > > actions=output:3 > > n_packets=255,n_bytes=24990,priority=2000,icmp,nw_src=10. > 0.0.2,nw_dst=10.0.0.3,icmp_type=8,icmp_code=0 > > actions=output:2 > > Are you sure that those are even the same flows? Somehow, the timeout > has disappeared entirely. > > What switch is this? > -- With regards, Quamar Niyaz
_______________________________________________ openflow-discuss mailing list openflow-discuss@lists.stanford.edu https://mailman.stanford.edu/mailman/listinfo/openflow-discuss
