roaming profile users get left in limbo (no login)

[amfony]

Hi Gang,

 

Ive got this problem, which ive always had to a degree, where roaming prof users will login and sometimes then be left at a post-login/pre-explorer starting limbo.

 

The problem is now it is happening to almost everyone and almost all the time, this sudden increase happened overnight practically.

 

Their only recourse is to CTRL + ALT + DEL and logoff and try again, which doesnt always work. But eventually will.

 

I am trying to do 2 things:

 

1) see if anyone can help? (the profile srv is a 2003 r2 super-specced box, client are XP sp3 and some SP2)

 

2) see if anyone can help me find the registry edits needed to enalbe Profile loading logging. I simply cannot find this anywhere after looking for about an hour.

 

Can anyone shed some light? OR have the same problem?

 

My first attempt to hack a fix was to check if explorer.exe exists if not start it in the login script, but this doesnt work as this issue is aparent before logon scripts are run.

 

Please help if you can.

[denon101]

ok

 

You can enable verbose logging by going to in to your workstation GPO. Then go Administrative Templates / System / Verbose vs Normal messages. Set this to enable.

 

You will now see when users logon they will get more messages in the loading personal settings box etc. Plus your log files will be far more detailed.

 

I have a similar problem but I have machines that hang when applying personal settings, but a reboot usually clears it.

 

Hope that helps

[srochford]

2) see if anyone can help me find the registry edits needed to enalbe Profile loading logging. I simply cannot find this anywhere after looking for about an hour.

 

 

I think what you want is called userenv debug logging - details on the MS web site

 

it feels like a server problem unless you've perhaps made some change to all the workstations (permissions tightened on c:\documents and settings? - unlikely because that would mean user X would never log on)

 

Dodgy network cabling or switch? Can you run wireshark? What you're looking for is excess traffic (and for this you really need to know what "normal" looks like; loopbacks on a switch can cause symptoms like this and you'll see masses of traffic. Lots of it is "normal" but just retrying and retrying)

 

Failing disc on the server? Anything logged in event viewer or disc management tools?

 

Run performance monitor on the server - what does the disc queue look like (should be <1 for normal working; anything more suggests the disc is stressed; then need to find what is causing that stress)