G
Guest
I have a site running w2k server, sql2000, and a bunch of scheduled tasks. every few weeks the registry reaches the maximum size ( was increased to 180MB !!). I was there late last week and it was only 16mb. Today got a call to say it had reached the limit again.
What seems to happen is that the user profile in use duplicates itself, so you end up with version .000, .001, .002, .003, and also several .bak copies of the same
They initially thought it was a problem with the build, so they started again on another box, and lo its started again
I have checked the sys & App logs but nothing appears obvious
They also lose all app settings / mappings after this and have to recreate them all, and not logout again as it all needs resetting.
They use Term Serv to remotely administer and the console is always logged in as the system admin. I have suggested using another account when using TS. but nothing obvious is wrong
I want to logout all users, remove all dupicated profiles and recreate the sys admin, but as this system is running constant data transfer for a monitoring & accounting system they are reluctant to give any downtime, even logging aout, in case of any knock on effects
any ideas ? cheer
paul
What seems to happen is that the user profile in use duplicates itself, so you end up with version .000, .001, .002, .003, and also several .bak copies of the same
They initially thought it was a problem with the build, so they started again on another box, and lo its started again
I have checked the sys & App logs but nothing appears obvious
They also lose all app settings / mappings after this and have to recreate them all, and not logout again as it all needs resetting.
They use Term Serv to remotely administer and the console is always logged in as the system admin. I have suggested using another account when using TS. but nothing obvious is wrong
I want to logout all users, remove all dupicated profiles and recreate the sys admin, but as this system is running constant data transfer for a monitoring & accounting system they are reluctant to give any downtime, even logging aout, in case of any knock on effects
any ideas ? cheer
paul