WDM 4.5.1: Jobs do not execute

  • This topic is empty.
Viewing 3 posts - 1 through 3 (of 3 total)
  • Author
    Posts
  • #534
    AvatarJoe
    Member
    • Total Post: 106
    • Legend in Own LunchBox
    • ★★★★★

    I’m having a problem with WDM 4.5.1 that I’m working with Wyse on. The software is starting to frustrate me, as no errors occur, things just are not working. Nothing in the event log, nothing in WDM Error logs — nothing.

    The problem is:
    A job is submitted to a client. It can be a PXE image push, a config pull, it doesn’t matter what.

    The job will enter the update manager que and sit there for about an hour. Sometimes I can kick off the job manually by submitting it then refreshing the device from the device manager.

    I’ve used MDTOOLS to clear the server and logs table, and that seemed to work for a little bit. But then, overnight, the problem started happening again.

    It seems like the jobs will only updated when the clients check in. They are all CE devices.

    I’m running a fresh version of 4.5.1 that had the DB installed to an MSDE instance then moved to a SQL Server 2000 Instance.

    I haven’t done much of anything to the install other than importing over 200 packages from our previous 4.4 install.

    It seems very similar tot he problems in this thread.

    Any ideas?

    #9943
    AvatarJoe
    Member
    • Total Post: 106
    • Legend in Own LunchBox
    • ★★★★★

    A little more information…

    I’ve confirmed that DEP is only running for Windows Essential Programs and Services.

    I bumped up logging and I see a few weird things out there. The first thing is an error that pops up once in a while that says “Dispatch DB Connect Failed.” Not quite sure what that means.

    The second one, and what I think is the problem, is that I see that when I submit a job, it logs an entry that says the device already has a job scheduled for it. That is not correct at all. I’ve gone so far as to clear the SystemLog and Server tables (per Wyse recommendation) and check the Command table to verify that nothing is defined here. I was told this was the table that it used to store the jobs. So, after verifying that no jobs at all where scheduled, I submitted a new job. The job was to push a default firmware to an s30. In the log it recorded that a job for that device already existed.

    If I clear out all jobs from the Update manager, clear the server and system log tables, then stop W3SVC, MSFTPSVC, and RAPPORT 4, Then start W3SVC, MSFTPSVC, RAPPROT4, and HSERERVERINIT, I can send one job to the machine. The job will finish successfully. Once it is done, if I try to send the same job to the machine, it does not work. It records a message to the logs saying that the machine already has a job scheduled for it.

    It’s almost as if the thin client isn’t checking in after the job and telling WDM that it has finished, even though the command table, as well as the Update Manager show that no jobs are pending.

    #10072
    AvatarJoe
    Member
    • Total Post: 106
    • Legend in Own LunchBox
    • ★★★★★

    I believe I have resolved this issue.

    It appears that some of the WDM time functions make use of the DB server, and some of the time functions make use of the WDM server. Since I had moved my DB to a different state, which was in a different time zone, everything was an hour off.

    After adjusting the time zone on the WDM server to match the DB server, the issues appear to be resolved. It’s been running for almost 24 hours without issues.

Viewing 3 posts - 1 through 3 (of 3 total)
  • You must be logged in to reply to this topic.