SE Agent crash happens when vnic that is meant for global vrf is not added first on Service engine.

Products

• VMware Avi Load Balancer

Issue/Introduction

In VMware Write Access Cloud or NSX Cloud with vnic that is meant for global vrf is not added first on Service engine.

If any new VS is getting placed on SE can cause se agent to crash and SE will go for reboot.

Stack trace: (Crash is available at following path "Administration--> Support --> Crash Reports)

Core was generated by `/opt/avi/bin/se_agent --log_dir=/var/lib/avi/log/glog --metrics_timer=60000 --r'.
Program terminated with signal SIGABRT, Aborted.
#0  0x00007f122cb4900b in raise () from /lib/x86_64-linux-gnu/libc.so.6
[Current thread is 1 (Thread 0x7f11fd7fa700 (LWP 1987))]
#0  0x00007f122cb4900b in raise () from /lib/x86_64-linux-gnu/libc.so.6
#1  0x00007f122cb28859 in abort () from /lib/x86_64-linux-gnu/libc.so.6
#2  0x00007f122cb9326e in ?? () from /lib/x86_64-linux-gnu/libc.so.6
#3  0x00007f122cb9b2fc in ?? () from /lib/x86_64-linux-gnu/libc.so.6

Environment

• VMware / NSX write access cloud
• AVI Load Balancer Version: 22.1.3, 22.1.4-2p1,22.1.4-2p2

Cause

In VMware Write Access Cloud or NSX Cloud, SE Agent crash may happen when there is no vNIC (because there is no VS) in global VRF and VS is present in other VRFs

vNIC meant for other VRF first move to global VRF and then to their target VRF. As there is no vNIC in global VRF, temporary vNIC addition to global VRF leads to global namespace creation and moving to other VRF leads to the namespace deletion.

When global VRF namespace creation and deletion happen within ~500 ms, it may lead to crash.

Resolution

Upgrade AVI Load Balancer version to 22.1.5, 30.1.1 and Above

We would recommend you to upgrade to latest AVI Load balancer version i.e 22.1.7-2px or 30.2.2-2px