vSAN does not support having scratch log on the vSAN Datastore as this can result in a hung host. For more information, see Redirecting system logs to a vSAN object causes an ESXi host lock up (2147541).

• For servers that support multiple controllers create a vmfs datastore on disk(s) attached to the secondary controller for storing persistent logging. Creating a persistent scratch location for ESXi 7.x/6.x/5.x/4.x (1033696)
• If the server only has one controller assign one disk for each host for boot and/or scratch partition. For more information see Best practices when using vSAN and non-vSAN disks with the same storage controller (2129050)
• Set up a Syslog server and have the logs stored here. See Configuring syslog on ESXi (2003322) to assist with setting this up.
  • If you set up a Syslog server and want to keep all logs in a central location, make sure to create subfolders per node in the main folder as the first node to start logging will create a lock on the log files preventing any of the other nodes from logging. Additionally, when the node holding the locks is rebooted the locks will migrate to another node that starts logging and now you have to separate nodes logging to the same log. Both scenarios will prevent support from troubleshooting the issue at hand in the environment fully.
• You can store coredumps on the SD boot media, but refrain from configuring the scratch partition here as the logs are write intensive and can cause the SD card to fail faster resulting in re-installation of ESXi
  • Minimum size of the SD card should be 16G
• For servers that support SATADOM you can configure your scratch partition here, but refrain from using for coredumps as the partition cannot be resized automatically as needed to store the core file.

Additional Information

For translated versions of this article, see: