* I have a workstation with the ZENworks 10.2.0 Adaptive Agent.

* I have servers running ZENworks 10.3.3.

* In ZCC, I tell it to deploy the 10.3.3 update to that workstation.

* On the workstation, it first deploys the 10.3.0a update, which is apparently a prerequisite for deploying 10.3.3. This update finishes as "Update Completed With Errors", because of the CASA hang I've described in another thread.

* The workstation reboots, and attempts to deploy the 10.3.3 update. This update fails as "One or more pre-requisite updates failed to install to this device".

* I take steps to attempt to fix the problem which caused the 10.3.0a update to partly fail.

* In ZCC, in the list of workstations for which the 10.3.3 update failed, I tell it "Redeploy update to workstations" for the affected workstation.

* On the workstation, it attempts to redeploy the 10.3.3 update. This update fails, because the prerequisite update still is not properly installed.

If I manually set up a deployment for 10.3.0a and point it at this workstation specifically, it re-tries the update, and in some cases succeeds normally. If I then attempt to redeploy the 10.3.3 update, it proceeds as normal.

However, it seems bizarre that ZENworks would attempt to redeploy the update which failed because of a missing prerequisite, rather than attempting to first redeploy the prerequisite - given that it did attempt to automatically deploy the prerequisite 10.3.0a in the course of deploying 10.3.3 in the first place. It also seems potentially odd that ZENworks would even attempt to proceed with the 10.3.3 update in the first place, given that 10.3.0a at least partly failed to deploy correctly.

Are these behaviors intended? Is there any way to configure or control either of them?

Is there any way to trigger the "one update right after another" behavior again, so that we can proceed with updating the partly-failed workstations without needing to manually select them all for deployment of 10.3.0a?