Some ESXi host transport nodes status are UNKNOWN.

Products

• VMware NSX

Issue/Introduction

• The host transport node status is displayed as "Unknown".
• Mgmt/CCP connections and tunnel status are fine.
• There are logs indicate that nsx-sha memory usage reached to the limit in vmkernel log.

2025-01-13T08:24:05.629Z cpu29:427152445)Admission failure in path: host/vim/vmvisor/nsx-sha:nsx-sha.428332292:uw.428332292
2025-01-13T08:24:05.629Z cpu29:427152445)UserWorld 'nsx-sha' with cmdline 'unknown'
2025-01-13T08:24:05.629Z cpu29:427152445)uw.428332292 (3810519913) extraMin/extraFromParent: 128/128, nsx-sha (3737352854) childEmin/eMinLimit: 56241/56320
2025-01-13T08:24:05.629Z cpu29:427152445)Admission failure in path: host/vim/vmvisor/nsx-sha:nsx-sha.428332292:uw.428332292
2025-01-13T08:24:05.629Z cpu29:427152445)UserWorld 'nsx-sha' with cmdline 'unknown'
2025-01-13T08:24:05.629Z cpu29:427152445)uw.428332292 (3810519913) extraMin/extraFromParent: 128/128, nsx-sha (3737352854) childEmin/eMinLimit: 56241/56320
2025-01-13T08:24:05.631Z cpu29:427152445)WARNING: LinuxThread: 424: nsx-sha: Error cloning thread: -28 (bad0081)
2025-01-13T08:24:10.636Z cpu19:427152444)Admission failure in path: host/vim/vmvisor/nsx-sha:nsx-sha.428332295:uw.428332295
2025-01-13T08:24:10.636Z cpu19:427152444)UserWorld 'nsx-sha' with cmdline 'unknown'
2025-01-13T08:24:10.636Z cpu19:427152444)uw.428332295 (3810519922) extraMin/extraFromParent: 128/128, nsx-sha (3737352854) childEmin/eMinLimit: 56241/56320
2025-01-13T08:24:10.636Z cpu19:427152444)Admission failure in path: host/vim/vmvisor/nsx-sha:nsx-sha.428332295:uw.428332295
2025-01-13T08:24:10.636Z cpu19:427152444)UserWorld 'nsx-sha' with cmdline 'unknown'
2025-01-13T08:24:10.636Z cpu19:427152444)uw.428332295 (3810519922) extraMin/extraFromParent: 128/128, nsx-sha (3737352854) childEmin/eMinLimit: 56241/56320
2025-01-13T08:24:10.638Z cpu19:427152444)WARNING: LinuxThread: 424: nsx-sha: Error cloning thread: -28 (bad0081)

Environment

VMware NSX

Cause

There was a memory leak issue in the process nsx-sha.

Resolution

This issue is resolved in NSX 4.2.0 .

A temporary workaround is to restart the nsx-sha with the following command on the hosts:

/etc/init.d/netopad restart