R
Rickamon
I have done extensive searching and have not found a thread on this
yet. I'm not sure if this is the appropriate group for this question
but I will start here with it...
I am a IT technician for a university that is in desperate need of
assistance. I've been battling a problem that has plagued our network
since the beginning of the academic school year. I created and deployed
the image that is currently on 80% of the campus lab computers. It was
deployed a week prior to the beginning of the semester with no problems
reported at all. The first week of class was a different story. We
received a large number of reports that many students were having
trouble logging on or logging off computers. Specifically, for those
that could not log on, after inputting their username and password they
saw only the desktop; no taskbar and no desktop icons. For those that
were logging off, again they would see the only the desktop but the log
off would not complete (I believe they could see the log off dialog box
as well but it would never disappear and never return to the "press
alt-ctrl-del to log on" dialog box). Here's the kicker, this does not
happen to every user and not on the same PCs. One user may have a
problem logging on to a particular computer but the next person who
tries to log into the same computer gets in with no problems.
Incidentally the first user who had trouble logging into said computer
can usually jump to another computer and log in fine (rules out account
issues). Sometimes a reboot will fix this problem but it doesn't
always.
Our initial feeling was that it might be something on my image that
could be causing the lock up, i.e. driver conflict, incorrect
permissions, etc... I have rebuilt the image at least 4 times with the
same problem occurring roughly 2 weeks after each deployment. Then we
caught wind of a possible cause for this problem. It was recently made
known to us that the current number of domain controllers on our
network may not be sufficient to handle the recent changes made by our
systems group. Specifically, they have migrated several services from
aging systems to our current DCs. After doing so they noticed high
hardware utilization on our 2 DCs which may cause the DCs to become so
busy that they cannot fullfil a log on or log off request. They have
recently set up a third DC and started distributing the work load to it
(although I do not know if this has been completed). I thought that our
log on/log off issues had cleared up. This morning I was informed that
we are starting to get reports of "lock ups". We are unsure just how
this DC issue figures into our problems, if at all.
We have been exploring the event logs to find some clue as to what's
going on. We sometimes see warnings concerning: user environments,
insufficient resources, and inability to contact a domain controller.
The user environtment errors were occuring on the previous build of the
image last year, so we have ruled that event as being insignificant.
The insufficient resouces error baffles us because it's just not the
case with these machines. Everytime we get a report of trouble we check
to see if the harddrive is full of user profiles. None of our machines
have ever been filled to capacity due to profiles. These machines come
with 2 gigs of ram and a 160GB harddrive so we have no idea why this
error is popping up (the bulk of the campus machines are Dell Optiplex
GX620 with the rest being different Optiplex models). We are still
looking into this as a possible source of our problem. It may be worth
noting that we do not use roaming profiles. As for the domain
controller error we really don't have anything for this. As we do not
have direct access to the DCs we can't really explore this error. Our
systems group is aware of this (seen it with their own eyes) but we
have yet to hear what could be causing this and how signficant this
error is.
Some of the things we have done to try to recetify this problem are as
follows:
-installed a profile cleaner (UPHcleaner) that will assist with the
computer releasing the registry after a user logs out as it should.
-manipulated group policy several times in an attempt to locate/fix the
problem.
-created images specific to each type of machine to rule out driver
conflicts
It's six weeks into the semester and we are no closer to isolating and
fixing this problem. With the problem being as inconsistant as it is
extreme frustration is setting in. Any help will be appreciated. If
more information is needed ask for it. I will be more than happy to
provide it.
Thanks
yet. I'm not sure if this is the appropriate group for this question
but I will start here with it...
I am a IT technician for a university that is in desperate need of
assistance. I've been battling a problem that has plagued our network
since the beginning of the academic school year. I created and deployed
the image that is currently on 80% of the campus lab computers. It was
deployed a week prior to the beginning of the semester with no problems
reported at all. The first week of class was a different story. We
received a large number of reports that many students were having
trouble logging on or logging off computers. Specifically, for those
that could not log on, after inputting their username and password they
saw only the desktop; no taskbar and no desktop icons. For those that
were logging off, again they would see the only the desktop but the log
off would not complete (I believe they could see the log off dialog box
as well but it would never disappear and never return to the "press
alt-ctrl-del to log on" dialog box). Here's the kicker, this does not
happen to every user and not on the same PCs. One user may have a
problem logging on to a particular computer but the next person who
tries to log into the same computer gets in with no problems.
Incidentally the first user who had trouble logging into said computer
can usually jump to another computer and log in fine (rules out account
issues). Sometimes a reboot will fix this problem but it doesn't
always.
Our initial feeling was that it might be something on my image that
could be causing the lock up, i.e. driver conflict, incorrect
permissions, etc... I have rebuilt the image at least 4 times with the
same problem occurring roughly 2 weeks after each deployment. Then we
caught wind of a possible cause for this problem. It was recently made
known to us that the current number of domain controllers on our
network may not be sufficient to handle the recent changes made by our
systems group. Specifically, they have migrated several services from
aging systems to our current DCs. After doing so they noticed high
hardware utilization on our 2 DCs which may cause the DCs to become so
busy that they cannot fullfil a log on or log off request. They have
recently set up a third DC and started distributing the work load to it
(although I do not know if this has been completed). I thought that our
log on/log off issues had cleared up. This morning I was informed that
we are starting to get reports of "lock ups". We are unsure just how
this DC issue figures into our problems, if at all.
We have been exploring the event logs to find some clue as to what's
going on. We sometimes see warnings concerning: user environments,
insufficient resources, and inability to contact a domain controller.
The user environtment errors were occuring on the previous build of the
image last year, so we have ruled that event as being insignificant.
The insufficient resouces error baffles us because it's just not the
case with these machines. Everytime we get a report of trouble we check
to see if the harddrive is full of user profiles. None of our machines
have ever been filled to capacity due to profiles. These machines come
with 2 gigs of ram and a 160GB harddrive so we have no idea why this
error is popping up (the bulk of the campus machines are Dell Optiplex
GX620 with the rest being different Optiplex models). We are still
looking into this as a possible source of our problem. It may be worth
noting that we do not use roaming profiles. As for the domain
controller error we really don't have anything for this. As we do not
have direct access to the DCs we can't really explore this error. Our
systems group is aware of this (seen it with their own eyes) but we
have yet to hear what could be causing this and how signficant this
error is.
Some of the things we have done to try to recetify this problem are as
follows:
-installed a profile cleaner (UPHcleaner) that will assist with the
computer releasing the registry after a user logs out as it should.
-manipulated group policy several times in an attempt to locate/fix the
problem.
-created images specific to each type of machine to rule out driver
conflicts
It's six weeks into the semester and we are no closer to isolating and
fixing this problem. With the problem being as inconsistant as it is
extreme frustration is setting in. Any help will be appreciated. If
more information is needed ask for it. I will be more than happy to
provide it.
Thanks