We have detected a core-dump in the backup RE of router.

It seems the core times coincides with a configuration change.

show system core-dumps

re0:

--------------------------------------------------------------------------

re1:

--------------------------------------------------------------------------

-rw-rw---- 1 root root 687831458 Feb 9 12:39 /var/core/re1/evo-pfemand.re.re1.13972.2026_02_09.12_39_02_CET.tar.gz

total files: 1

23 2026-02-09 12:41:14 CET by nemeth1ernoa16 via cli commit synchronize

24 2026-02-09 12:39:02 CET by nemeth1ernoa16 via cli commit synchronize

25 2026-02-09 08:14:57 CET by nemeth1ernoa16 via cli commit synchronize

show system rollback compare 25 24

[edit interfaces et-0/0/20]

+ speed 1g;

+ mtu 9114;

+ ether-options {

+ auto-negotiation;

+ }

+ unit 725 {

+ vlan-id 725;

+ family inet {

+ mtu 1500;

+ address a.b.c.d/24;

+ }

+ }

This issue is addressed via internal PR 1911792.

Root Cause:

Stale Weak pointers were stored in redundant data structure.

2 problems are seen here,

> Weak pointer map is not getting cleaned when IFL is getting deleted

> While updating shared memory we are using both shared and weak pointer maps simultanously

Fix Details:

Make sure we are using shared pointer directly while IFL shared memory update

Observations:

Issue seen on standby RE

Test box reboot/PFE restart

Trigger:

Test IFL/IFD modify/creation/delete