NSX-ALB Service Engine crash with Stack trace error - frag6_input

Products

• VMware NSX

Issue/Introduction

Symptoms:

NSX-ALB Service engines go into a crash loop.
The stack trace would look like:

[New LWP 1989]
[New LWP 1992]
[New LWP 1994]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Core was generated by `se_dp: worker process 0:                          '.
Program terminated with signal SIGSEGV, Segmentation fault.
#0  0x0000560d16303883 in frag6_input (mp=mp@entry=0x7ffee1618378, offp=<optimized out>, proto=<optimized out>) at /home/aviuser/workspace/22.1.1-ci-build/service_engine/ipstk/ipstk-bsd/netinet6/frag6.c:541
----snip----

Please note that the memory location will be different. The above example is just for reference.

Cause

A specific variable getting uninitialized due to compiler optimization causing a crash.

Resolution

This issue has been resolved in 21.1.6 and 22.1.3.

Workaround:

If IPv6 is not being used in the environment - Login to the controller shell and run the following commands -

configure serviceenginegroup <se_group_name>
deactivate_ipv6_discovery
save

Additional Information

Impact/Risks:

Data traffic could be impacted due to service engine continuously crashing.