Recently I started to get a lot of warning messages like this in the Altiris log:
Log File Name: C:\Program Files\Altiris\Notification Server\nsLogs\a.log
Priority: 2
Date: 6/25/2008 10:30:16 AM
Tick Count: 8747765
Host Name: XXXXXXX
Process: w3wp.exe (1928)
Thread ID: XXXX
Module: AltirisNativeHelper.dll
Source: Altiris.ClientTask.Web.GetClientTaskServers.LocateServersBasedOnSubnet
Description: Cannot find any TCPIP information for machine XXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXX
The message appears every 5 minute after collection update task.
Since there is no computer name, only a GUID. I assumed that the computer was deleted from the database (probably by maintenance task) and some leftover caused the NS to think that it still exists.
In that case, I thought, I can clean the leftovers from the DB and every thing will be just fine. To filter out the entries with the GUID I add to the table view this line:
WHERE (ResourceGuid='xxxxxx-x THE GUID x-xxxxx-xxxx'}
(sometimes I had to use "_ResourceGuid" or just "Guid")
I tried to delete some entries in different tables that contain the same guid (don't forget to backup first!), but I couldn't find anywhere in the DB the cause to the error. But what I did managed to find in table dbo.ResourceKeyChanged
is the real computer name that was linked to that guid.
Fortunately the computer was online and the Altiris agent still exist on it. All I had to do is to push the agent again, and since it was already installed it registered to the NS server with the same GUID.
Now the warning messages stop appearing!
At that point I removed the agent normally and the computer was deleted from DB correctly.
At the bottom line, this is the best solution that I could get since I really didn't like the idea of deleting DB entries. You never know how this kind of action will impact the databse integrity. So the best thing is to avoid it if you can.