Scenario:
A Citrix Administrator updates all of the machines within a Delivery Group. After the update, an
application stops working. The IT manager tells the administrator to revert all machines to the
previous image. How should the administrator revert all machines to the previous image?
A.
By selecting ‘Rollback machine update’ for the Delivery Group
B.
By deleting the snapshot associated with the template virtual machine
C.
By selecting ‘Rollback machine update’ for the Session Machine Catalog
D.
By deleting the difference disks of the virtual machines within the Delivery Group
You use to rollback feature from the Machine Catalog.
Quote from https://discussions.citrix.com/topic/334835-update-and-rollback-catalog/
“Quick overview of how MCS works.
1) Admin creates VM and installs required software on it
2) A snapshot is taken of the VM (either by the admin or automatically by Studio)
3) The snapshot is supplied to MCS which takes a full copy of the disk data and flattens any delta disk hierarchy. Once this is complete the snapshot taken in 2 is not used by XD in any way and can be deleted if desired.
4) VM are created and given a disk which is a delta from the disk created in 3.
5) Admin updates the original VM with some changes (service pack, new software, etc.)
6) A new snapshot is taken
7) 3 is repeated to get a new base disk
8) Vms are updated to point at the new disk as they boot next time
9) Once all VMs are updated, the old base disk is removed
Now, Studio has an option to “rollback”, what this actually does is select a snapshot that was previously used and performs operation 3 again. Where this can go wrong is if previous snapshots have been removed and so the hierarchic path to the snapshot has changed. This doesn’t mean that rollback is impossible it just needs to be performed explicitly by selecting the correct snapshot and updating to it (even if update actually means revert to older). This is why it is a good idea to create the snapshots of the master VM yourself rather than allowing Studio to create it on your behalf as it means you can give it a descriptive name which simplifies determining which snapshot you want. Obviously, if you’ve deleted the snapshot you want to go back to then there isn’t a lot that can be done, so be sure that you aren’t going to want the snapshot again before deleting them.
As the base-disks that MCS uses are full copies of the master disk they can be quite large and so you don’t want to have too many of them hanging around. As I mentioned above, they will be removed automatically once no VMs are using them, to check this you can use the Search node in Studio and select the filter option “Pending Updates”, any machines listed in this search are holding old disk images into the system. Booting them (or rebooting if they are already on), will apply the update and release the image which will be cleaned up by a background housekeeping task which periodically checks the state of old disks.”
1
0