This is expected behaviour. HA is about availability and you do not want to disable this.
Your underlaing hypervisor crached or similar like this and VMs ware restarted on the other nodes.
You can disable HA and keep DRS running but once your ESXi crashes your VMs wont be restarted and be unaccessible until the affected ESXi is back online. DRS will still migrate VMs across the cluster.
Please double check your cluster settings for HA, is it enabled with VM monitoring?
Other thing worth checking would be, is any of the host facing failures, May be host is restarted due to some technical issues, that might trigger host monitoring of HA to initiate restart of VMs.
It's not advisable to disable vSphere HA & DRS on your cluster.