ESXi syslog collection to a NFS datastore stops if the host is rebooted

Products

• VMware vSphere ESXi

Issue/Introduction

Symptoms:

• ESXi syslog collection to a NFS datastore stops if the host is rebooted.
• This issue occurs when the ESXi syslog is configured to use a local NFS datastore ( Syslog.global.logDir).
• After the host reboots, the .vmsyslogd.err log file (located at /var/log/) contains entries similar to:
  
  YYYY-MM-DDT08:57:37.262Z vmsyslog : CRITICAL] vmsyslogd daemon starting (2403)
  YYYY-MM-DDT08:58:43.134Z vmsyslog.config : ERROR ] Config error: Invalid config: could not create log directory
  YYYY-MM-DDT08:58:43.135Z vmsyslog : ERROR ] Exiting with parse failure
  YYYY-MM-DDT08:58:44.288Z vmsyslog.main : ERROR ] reloading (2403)
  YYYY-MM-DDT08:58:44.290Z vmsyslog.config : ERROR ] Config error: Invalid config: could not create log directory
  YYYY-MM-DDT08:58:44.290Z vmsyslog.config : ERROR ] Config error: Invalid config: invalid log directory: /vmfs/volumes/d126f0a0-4a8846fe/syslogs/esx001; will instead use default: /scratch/log.

Environment

VMware vSphere ESXi 5.0
VMware vSphere ESXi 5.1

Resolution

This is a known issue affecting ESXi 5.0 and 5.1. Currently, there is no resolution.

To work around this issue, perform one of these options:

• Manually restart syslog collection after the reboot by running this command on ESXi:
  
  esxcli system syslog reload.
  
  
• Reconfigure syslog to use a VMFS datastore instead of a NFS datastore.
• Send syslogs over a network to a remote syslog collector ( Syslog.global.logHost). For example, you can use the syslog collector bundled with vSphere 5.x.

Additional Information

To be alerted when this document is updated, click the Subscribe to Article link in the Actions boxConfiguring syslog on ESXi