This article explains the meaning and implications of segmentation fault (segfault) logs in Juniper devices, specifically focusing on errors related to libjnh.so. It provides insights into interpreting such logs and troubleshooting potential issues that arise from them.

After performing a configuration commit on the device, the issue manifests almost immediately, leading to significant system instability. The symptoms observed include:

• Core Dumps:

Shortly after the configuration commit, the device generates core dumps.

• Port Failures:

In conjunction with the core dumps, the device experiences port failures.

• FPC went down:

We were able to see the fpc going down.

The logs show a segmentation fault (segfault) on FPC4 where the Direct-root process (PID 2425) crashed due to an invalid memory access attempt. The fault occurred within the libjnh.so shared library, which is involved in network-related functions in Junos.

The error code "error 4" indicates a memory access violation, typically caused by accessing an invalid memory address.

Mar  6 14:43:41.724 2025  router mgd[96548]: %INTERACT-6-UI_COMMIT_PROGRESS: Commit operation in progress: signaling 'Alarm control process', pid 28719, signal 30, status 0 with notification errors enabled
Mar  6 14:43:41.831 2025  router mgd[96548]: %INTERACT-6-UI_NETCONF_CMD: User 'ansible' used NETCONF client to run command 'unlock-configuration'
Mar  6 14:43:41.000 2025  router : %PFE-6: fpc4 kern.info kernel:Direct-root[2425]: segfault at 0 ip 00007f08a8d40cac sp 00007f08689c47f0 error 4 in libjnh.so[7f08a8848000+70f000]
Mar  6 14:43:41.878 2025  router : %USER-1: fpc4 1 2025-03-06T14:43:41.877928+00:00 cc-ss3-ams5-j960-re0-fpc4  - - [timeQuality tzKnown="1" isSynced="0"] Coredump started for cc-ss3-ams5-j960-re0-fpc4-aftd-trio
Mar  6 14:43:41.882 2025  router kernel: %CONSOLE-6: Mar  6 14:43:41.878 2025  cc-ss3-ams5-j960-re0 fpc4 1 2025-03-06T14:43:41.877928+00:00 cc-ss3-ams5-j960-re0-fpc4  - - [timeQuality tzKnown="1" isSynced="0"] Coredump started for cc-ss3-ams5-j960-re0-fpc4-aftd-trio

A core dump was triggered due to the segfault, which is essential for further analysis to determine the root cause. It could be caused by software bugs, resource issues, or hardware faults.

The crash seemed to be related to a recent configuration change involving firewall policer settings, as indicated by the backtrace in the core dump for this case.

Next Action:

To determine the exact Root Cause Analysis (RCA), it is essential to decode the core dump. These dumps will help identify specific processes or system components that encountered critical errors or crashes, which could be due to improper configurations, system overloads, or software bugs.

By analyzing the core dump, we can gain insights into whether the issue is related to software bugs, resource exhaustion, or potential incompatibilities introduced by recent configuration changes. Decoding the core dump will be the key to pinpointing the exact cause of the failure and formulating the necessary resolution steps.

Contact Juniper Technical Assistance Center (JTAC) for further assistance.