vSAN iSCSI target service interrupted if ESXi host agents are restarted

Products

• VMware vSAN

Issue/Introduction

Symptoms:

• vSAN iSCSI target service IO could be interrupted and the service cannot be started when restarting the ESXi host agents.
• In the /var/run/log/vitd.log, you see similar to:

2020-06-16T04:50:25Z vitd: [etc/init.d/vitd] SIGKILL has been sent to vitsafehd
2020-06-16T05:42:38Z vitd: [etc/init.d/vitd] vitd is not running
2020-06-16T05:42:38Z vitd: [etc/init.d/vitd] vitsafehd is not running
2020-06-16T05:58:31Z vitd: [etc/init.d/vitd] vitd is not running
2020-06-16T05:58:31Z vitd: [etc/init.d/vitd] vitsafehd is not running

Environment

VMware vSAN 8.x

VMware vSAN 7.x
VMware vSAN 6.x

Cause

When the ESXi host agents are restarted, it will shut down the vSAN iSCSI vitd/vitsafehd as well as the watchdog daemon. As these are vSAN processes the ESXi host agents won't restart these services as part of a management agents restart.

Resolution

Be mindful of the ESXi host agents that need to be restarted during troubleshooting and only restart the host services that are absolutely needed to resolve the issue at hand.

Workaround:
If ESXi host agents have been restarted and vSAN iSCSI is not functioning as intended to work around this issue, run this command on the affected ESXi host(s):
/etc/init.d/vitd io_reset

Additional Information

vSAN Health Service - vSAN iSCSI target service – Service runtime status (2147603)

Impact/Risks:
vSAN iSCSI I/O interrupted.