Logging to vmkernel.log has stopped in ESXi
search cancel

Logging to vmkernel.log has stopped in ESXi

book

Article ID: 414052

calendar_today

Updated On:

Products

VMware vSphere ESXi

Issue/Introduction

There are cases where output to vmkernel.log may stop due to IO errors. As a result, it may not be possible to track down the details of recent network or storage issues from it.

Environment

  • ESXi 8.x
  • ESXi 7.x

Resolution

In some circumstances, restarting vmsyslogd will cause past vmkernel logs to be output again.

 1. Restarting vmsyslogd

ESXi 7.x

See ESXi host logs not updating despite no errors observed

ESXi 8.x

 watchdog -k vmsyslogd

/usr/lib/vmware/vmsyslog/bin/vmsyslogd

2. Open vmkernel.log to determine if any cached logging has been flushed to the file

Additional Information

Unlike userworld logs, VMkernel logs are written to a special char device and are saved for a certain period of time until they are overwritten.
By restarting vmsyslogd, you can output the log contents remaining on that char device to the vmkernel.log.
If logs related to the most recent problem remain on that char device, restarting vmsyslogd will cause them to be re-output to the vmkernel.log.

Note: If you are forwarding logs and monitoring them, please be aware that past logs will be output again, which may result in the detection of problems that occurred in the past.