Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 2054102

Summary: [OCP 4.8][OVN] packet loss when communicating between pods
Product: OpenShift Container Platform Reporter: Aaron Park <aapark>
Component: NetworkingAssignee: Andreas Karis <akaris>
Networking sub component: ovn-kubernetes QA Contact: Anurag saxena <anusaxen>
Status: CLOSED DUPLICATE Docs Contact:
Severity: urgent    
Priority: medium CC: anbhat, bzvonar, dcbw, ffernand, hdo, jinjli, mcambria, openshift-bugs-escalate, tidawson, trozet
Version: 4.8   
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2022-03-11 17:53:42 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Aaron Park 2022-02-14 07:43:04 UTC 
Description of problem:

Customers are performing validations on clusters for Verizon operators.
They load the Pods with heavy traffic, and packet loss occurs when communicating between Pods.


Version-Release number of selected component (if applicable):

- OCP 4.8.24(OVN + IPv6)
- network
Network : 
spec:
  clusterNetwork:
  - cidr: fd01::/48
    hostPrefix: 64
  externalIP:
    policy: {}
  networkType: OVNKubernetes
  serviceNetwork:
  - fd02::/112


How reproducible:

- Packet loss occurs during communication between pods
UPP Pod -> UIP Pod

- Packets sent to UPP -> UIP are caught as dumps,
No packets were received in the UIP dump.

At UPP side:
08:46:06.602831 IP6 (class 0xe0, flowlabel 0x23553, hlim 64, next-header UDP (17) payload length: 264) fd01-0-0-5--24c5.aupf9901-uip0.cnf-aupf.svc.cluster.local.27035 > aupf9901-upp0-0.aupf9901-upp0.cnf-aupf.svc.cluster.local.27120: [udp sum ok] UDP, length 256

08:46:06.604583 IP6 (class 0xe0, flowlabel 0x67f7e, hlim 64, next-header UDP (17) payload length: 44) aupf9901-upp0-0.aupf9901-upp0.cnf-aupf.svc.cluster.local.27120 > aupf9901-uip0.cnf-aupf.svc.cluster.local.27035: [bad udp cksum 0xaeac -> 0xb07b!] UDP, length 36 <---- UIP can't receive this msg

Steps to Reproduce:
1.
2.
3.

Actual results:

No packets were received in the UIP dump.

Expected results:

Communication should be done without packet loss.

Additional info:

- The customer has attached the data below.
  - worker node sosreport
  - must-gather
  - pcap for each pods
Comment 7 Andreas Karis 2022-03-11 17:53:42 UTC 
I will go ahead and close this out as a duplicate of 2018930. Let me know if you don't agree. Thanks

*** This bug has been marked as a duplicate of bug 2018930 ***