You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Mar 8, 2023. It is now read-only.
Our standard VM instantiation leverages multiple vHBAs to get multiple write-queues; it's a standard perf tweak for better offload, which then allows the hypervisor more latitude in write ordering with more writes in-flight.
+1
We pre-populate all of our VM templates with 4 pvscsi controllers. (3 of which don't have hard drives) This is due to the 'feature' of VMWare BIOS that the 4th peripheral device added to a VM get's a lower hardware ID than the first. So if you have a VM with 3 SCSI controllers, 3 disk and a NIC; then add a 4th SCSI controller, what was according to BIOS controller 0, 1, 2 is now 1, 2, 3, 0, making SCSI IDs seen by the OS different. On next boot if, or immediate if done live, Windows sees disk 0 that used to be 0:0 is now 1:0, disk 2 is now 2:0, disk 3 is now 3:0, sets disk 1 to offline, reattaches disk 0 and disk 3 with the new scsi ID and adds disk 4. Disk 1 must be brought back online manually.
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
After the successful fix for #68 , I'd like to talk about a feature to manage more than one iscsi controller as well:
Our standard VM instantiation leverages multiple vHBAs to get multiple write-queues; it's a standard perf tweak for better offload, which then allows the hypervisor more latitude in write ordering with more writes in-flight.
As mentioned under #68 (comment 530072308).
The text was updated successfully, but these errors were encountered: