Problem with some Vista machines on boot

[bhearsum]

Thanks - I'll give that a try today.

[bhearsum]

Unfortunately, making that change did not affect anything - OPSI continued to block while it installed.

More pressingly, I'd like to figure out why opsiclientd is failing to start. Do you have any advice on how to get more information on that? There wasn't anything useful in the system logs.

[j.schneider]

Are you sure opsiclientd is not running?
Can you start it manually?

Code: Alles auswählen

net start opsiclientd

You can increase the log level in opsiclientd.log.

[bhearsum]

Are you sure opsiclientd is not running?
Can you start it manually?

Code: Alles auswählen

net start opsiclientd

You can increase the log level in opsiclientd.log.

Sometimes it starts, sometimes it doesn't. We get the dialog described in my initial post when it doesn't start, and then we have to reboot the machine - so there's no way to inspect the machine when it happens. How do I turn up the log level for opsiclientd.log?

[bhearsum]

Another thing I just realized: I didn't have this problem in an earlier tests. It wasn't until I upgraded to preloginloader 3.4-27 that I had issues. Is it possible there's a bug in that package that causes opsiclientd to not start on Vista sometimes?

[j.schneider]

You can set the log level in opsiclientd.conf:

Code: Alles auswählen

[global]
# Set the log (verbosity) level
# (0 <= log level <= 9)
# 0: nothing, 1: critical, 2: errors, 3: warnings, 4: notices
# 5: infos, 6: debug messages, 7: more debug messages, 9: passwords
log_level = 6

[j.schneider]

Could you please try if the new version http://download.uib.de/opsi3.4/testing/ ... .4-29.opsi fixes the problem ?

[bhearsum]

Thanks for working with me on this, I'll try the new preloginloader package.

[j.schneider]

Sorry, just found a bug in the new version.
The next version is ready for testing:
http://download.uib.de/opsi3.4/testing/ ... .4-30.opsi

[bhearsum]

I haven't gotten a chance to test the new preloginloader yet, but I think it's unrelated to our issues. I noticed that the OPSI server had very poor disk I/O performance, 3MB/sec or so, and other performance issues. By improving the load on the server we've had many fewer slaves fail.