One of my area schools was having strange problems - some users could not log in, some users got logged in but could not access their redirected home folders, and some users were prompted multiple times while trying to open Outlook.
I reotely checked a few things on the server and it appeared the Active Directory, File Replication, and Exchange systems had gone a bit haywire (the word I usually use here is "wonky").
Here is a timeline of what I did:
9:20a - Searched for steps to test AD database integrity and repair. Found the following:
I reotely checked a few things on the server and it appeared the Active Directory, File Replication, and Exchange systems had gone a bit haywire (the word I usually use here is "wonky").
Here is a timeline of what I did:
9:20a - Searched for steps to test AD database integrity and repair. Found the following:
Boot the server(s) to Directory Service Recovery Mode (DSRM)
In Windows\System32, run NTDSUTIL
Type FILES, hit ENTER
Type INTEGRITY, hit ENTER
If needed, RECOVER
Type QUIT to get back one menu
Type SEMANTIC DATABASE ANALYSIS
Type GO FIXUP
If needed, use COMPACT TO to defrag
In Windows\System32, run NTDSUTIL
Type FILES, hit ENTER
Type INTEGRITY, hit ENTER
If needed, RECOVER
Type QUIT to get back one menu
Type SEMANTIC DATABASE ANALYSIS
Type GO FIXUP
If needed, use COMPACT TO to defrag
9:28a - Shut down DC2 for DSRM
9:28a - While waiting for the server to shut down, I looked up info for checking/repairing Exchange. When I booted to DSRM, the integrity check came back normal for AD, which was not the same result I had received while remoted in. That is not really surprising. I found a registry entry for "WaitForNetwork" and added that for good measure. I also issued a "DFSUTIL /PurgeMupCache" command.
10:05a - Rebooted the server
10:10a - Checked DC2 and found a LOT of ENV ERRORS in the Event Log. i noticed the time was way off from the other server so I reset the time to match DC. The server actually passed the integrity tests and Semanitc DB Analysis. I added the "WaitForNetwork" registry entry and issues the Mup-clearing command.
10:40a - Rebooted DC2
10:30a (while tests ran on DC2) - Started a defrag on the Exchange Store.
Stop the Information Store
Use ESEUTIL to make sure shutdown was clean
Set the defrag command to create new files, just in case.
Use ESEUTIL to make sure shutdown was clean
Set the defrag command to create new files, just in case.
10:35a - For grins, I ran GPUPDATE on each server and a nearby workstation.
10:37a - File Replication Services reported that FRS "is no longer preventing the computer (DC2) from becoming a Domain Controller." I took that as VERY good news.
On DC2, I reinstalled "Help & Support" per MSDN article (KB937055) since "Help Service" was not installed.
11:40a - Defrag of PRIV1.EDB was at 68% and I had not started PUB1.EDB yet! Yikes!
11:48a - PRIV1 done after 4223 seconds. Ran the defrag on PUB1, and it completed in 9 seconds.
I renamed the PUB1/PRIV1 files to something else, then renamed the temp PRIV/PUB files to PRIV1/PUB1 respectively. I then moved the old files to another server for safekeeping. Will delete those later as long as everything works!
We tested several users, specifically those having various troubles with logins, documents, etc. Everything checked out fine!
No comments:
Post a Comment