We are currently running ZCM 10.0.3 that was updated from 10.0.1 via 10.0.2 on OES2 Linux. We have only 118 workstations active in the system, patch management, inventory and asset management are active.

Now we have had issues since day one, most were supposed to resolved in the 10.0.3 update. Some were, some were not. I have confirmed the system thinks it completed the 10.0.3 update successfully. However after the update I found that the agents at http://<ZCMSERVER>/zenworks-setup were not updated. A simple command line I found in the forums resolved that. However that got me to thinking if the agents did not update correctly what else did not?

Let me give you a rundown on some of our issues:

1. Policies are hit or miss, they fail as often as they work. Failures are impossible to troubleshoot. The server will indicate a policy error and when I check the details on the policy error only to discover it was a couple of days ago. I go check the station and it says all is well. This can even occur in just a couple of hours. The only way I find a real policy error is when the user complains abut something policy related not working. Sometimes on the station I click the "Click for details" on a policy status of failed to get a policy applied successfully message.

2. Bundles are hit or miss but seem to work the best of everything.

3. Hot list and workstation errors are worthless, there are too many errors. I cleared all messages system wide and within 36 hours some stations were up over 1500 errors. A lot of them are "ServiceExistsSysReqHandler threw an exception....A system shutdown is in progress." Which really isn't really an error, the system is shutting down. Of 118 workstations in the system 111 are in the hot list. Policy/Bundle errors never seem to clear even when I fix the issue and refresh the station.

4. The server will hit 100% utilization and sit there for days, until I reboot it. Java and dbsrv10 are the offending tasks. I no longer wait to reboot more than an hour or two now as everything else seems to slow to a crawl. Imagine having to re-boot a Linux box about once a week....

5. On the workstation side nzrSignaller.exe and rmenf.exe crash often (since 10.0.3 update). This kills all ZCM functions until the station is re-booted.

We have been taking the issues one at a time and trying to solve them. With the possibility of the previous updates not working correctly we may never get to the bottom of some issues and it might be time to try a sledge hammer approach. I was thinking we could build another server with the 10.0.3 media then add it to the zone and designate it primary, replicate all the content and the database to it, (not sure about how to do the database). Then kill the one we have. Then we *know* all the files are updated correctly. After which we can "refresh" the workstations agents from the known good server somehow.

Now I am sure it is not this easy, I probably will break something else if I try. So does anyone else think this might work? and if so might it help solve some of our issues? Any gotcha's to be concerned about?

Sorry for the long post but I feel like I am playing wack-a-mole with ZCM and I am not getting many hits.

Any and all suggestions/comments welcome.

Thanks for listening,

Yesterday it worked.
Today it is not working.
Windows is like that.