[NETVIRT-811] CSIT Sporadic failures - Arp learning suite - ping to learned FIB entry failing Created: 26/Jul/17  Updated: 03/May/18  Resolved: 18/Aug/17

Status: Resolved
Project: netvirt
Component/s: General
Affects Version/s: Carbon
Fix Version/s: None

Type: Bug
Reporter: Jamo Luhrsen Assignee: Aswin Suryanarayanan
Resolution: Cannot Reproduce Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Operating System: All
Platform: All


External issue ID: 8895

 Description   

https://logs.opendaylight.org/releng/jenkins092/netvirt-csit-1node-openstack-ocata-upstream-stateful-snat-conntrack-carbon/75/log.html.gz#s1-s7-t2



 Comments   
Comment by Jamo Luhrsen [ 31/Jul/17 ]

https://logs.opendaylight.org/releng/jenkins092/netvirt-csit-1node-openstack-newton-nodl-v2-upstream-stateful-snat-conntrack-carbon/122/log.html.gz#s1-s7-t2-k31

Comment by Vivekanandan Narasimhan [ 01/Aug/17 ]

Update on NETVIRT-811.

Vivek and I analyzed the CSIT logs, and it looks like the Flow Cleanup phase for External Network Test is not cleaning up all the flows.

From the below link:
https://logs.opendaylight.org/releng/jenkins092/netvirt-csit-1node-openstack-newton-nodl-v2-upstream-stateful-snat-conntrack-carbon/122/log.html.gz#s1-s7

it looks like NAT related flows are not getting cleaned up.

Secifically regarding NETVIRT-811, a stale NAT conntrack flow is present in the table-36 entry, that is failing ping in its reply path, i.e,

cookie=0x8000006, duration=203.299s, table=36, n_packets=5, n_bytes=331, priority=10,ip,tun_id=0x186a9 actions=set_field:0x30d52->metadata,ct(table=46,zone=5003,nat) IN COMPUTE_1 is the stale flow, which is not allowing the packet to reach cookie=0x90186a9, duration=144.439s, table=36, n_packets=0, n_bytes=0, priority=5,tun_id=0x186a9 actions=group:150025.

@Aswin,
Could you please help us in finding out why the conntrack flow was not cleaned up properly?

Thanks,
Kiran

Comment by Jamo Luhrsen [ 01/Aug/17 ]

https://logs.opendaylight.org/releng/jenkins092/netvirt-csit-1node-openstack-ocata-upstream-stateful-snat-conntrack-carbon/82/log.html.gz#s1-s7-t2-k31

Comment by Aswin Suryanarayanan [ 17/Aug/17 ]

I tried multiple times locally(in one node and two node setup). But did see an instance where this flow is not getting cleaned up. Also this failure does not happen last 15 run. Is it safe for us to assume this may have got fixed?

Comment by Jamo Luhrsen [ 18/Aug/17 ]

(In reply to Aswin Suryanarayanan from comment #4)
> I tried multiple times locally(in one node and two node setup). But did see
> an instance where this flow is not getting cleaned up. Also this failure
> does not happen last 15 run. Is it safe for us to assume this may have got
> fixed?

yeah, lets close this and I'll re-open if it shows back up. CSIT is
passing pretty reliably now so I think a lot of the sporadic bugs
(like this) have been fixed.

Generated at Wed Feb 07 20:22:32 UTC 2024 using Jira 8.20.10#820010-sha1:ace47f9899e9ee25d7157d59aa17ab06aee30d3d.