- Jun 30, 2004
- 16,336
- 1,890
- 126
So I had my new (home) server built, setup and running with 2012 R2 Essentials. AV software from AVAST installed; StableBit DrivePool installed; Serviio Media Server installed; SyncBack Se installed; backups for all client machines started and occurring successfully; users and machines added -- all checking out as "Kosher."
My backup solution in a nutshell involved using Windows Backup exclusively for the server OS disk, planning another backup of important data with SyncBack SE to a different disk and volume.
I decided to change out smaller drives in the drive pool with larger enterprise-class disks. (See the Newegg sale on "refurbished" 3K7000 Hitachi 3TB drives). Of course, I went through the process in Stablebit to drop a drive at a time from the pool and successively add back the replacement drives. Shouldn't be a problem -- should it?
Soon thereafter, I booted up the server and attempted to open the dashboard. It took about ten minutes for the dashboard to run! I clicked on the "Users" tab. Suddenly, all my users are just -- gone! I clicked on "Devices." All the client computers had disappeared!
Controlling my panic and inclination to reinstall the OS all over again with the software, the hard-to-find drivers -- I started poking around the web to see who else had been having this sort of trouble. Easy to find -- there were lots of folks.
Among the things they mentioned doing just before the symptoms occurred was new disk installation. What could it be? Why hasn't MS fixed it? How can I get back my dashboard?
Event log symptoms included failure to start WSSMgmtEng service. It would be in a "stopped" state, and attempting to start it would show "Starting" and then quickly flip back to "Stopped."
MS Techies tried proposing complex solutions, which many attempted, and few if any reporting success. In one thread, the posts just kept coming after one poster noted a simple way of fixing the problem. It was as though nobody noticed his comment.
What did he do? He executed the command " wbadmin delete catalog" from an elevated command window or Powershell. This deleted what apparently was corrupted catalog files for the server's own backup.
One could make some vague but sensible guesses as to why this had anything to do with the dashboard display of "users" and "devices." But people were wringing their hands and pulling their hair.
Apparently, adding a new disk can confuse the backup service in identifying the previously-configured backup drive.
Some advice for the server OSes suggests using a detachable or external USB drive for the server's own backup. I use a hot-swap drive in a 5.25" bay, which I leave connected most of the time, since backups are scheduled and automated.
But we all need to swap disks in and out.
What sort of procedure would guarantee this doesn't happen again?
I'd think one should probably remove the server backup configuration and disconnect the disk before adding new hardware, and then re-create the backup.
Certainly, someone who works in a data-center managing enterprise servers might have a clear idea of what is done routinely in those situations.
But I just had to post this thread with its questions. I might have completely missed the shortcut fix for it, and then reinstalled the OS and software from scratch.
Do people who work in data-centers managing servers ever thump their chests?
My backup solution in a nutshell involved using Windows Backup exclusively for the server OS disk, planning another backup of important data with SyncBack SE to a different disk and volume.
I decided to change out smaller drives in the drive pool with larger enterprise-class disks. (See the Newegg sale on "refurbished" 3K7000 Hitachi 3TB drives). Of course, I went through the process in Stablebit to drop a drive at a time from the pool and successively add back the replacement drives. Shouldn't be a problem -- should it?
Soon thereafter, I booted up the server and attempted to open the dashboard. It took about ten minutes for the dashboard to run! I clicked on the "Users" tab. Suddenly, all my users are just -- gone! I clicked on "Devices." All the client computers had disappeared!
Controlling my panic and inclination to reinstall the OS all over again with the software, the hard-to-find drivers -- I started poking around the web to see who else had been having this sort of trouble. Easy to find -- there were lots of folks.
Among the things they mentioned doing just before the symptoms occurred was new disk installation. What could it be? Why hasn't MS fixed it? How can I get back my dashboard?
Event log symptoms included failure to start WSSMgmtEng service. It would be in a "stopped" state, and attempting to start it would show "Starting" and then quickly flip back to "Stopped."
MS Techies tried proposing complex solutions, which many attempted, and few if any reporting success. In one thread, the posts just kept coming after one poster noted a simple way of fixing the problem. It was as though nobody noticed his comment.
What did he do? He executed the command " wbadmin delete catalog" from an elevated command window or Powershell. This deleted what apparently was corrupted catalog files for the server's own backup.
One could make some vague but sensible guesses as to why this had anything to do with the dashboard display of "users" and "devices." But people were wringing their hands and pulling their hair.
Apparently, adding a new disk can confuse the backup service in identifying the previously-configured backup drive.
Some advice for the server OSes suggests using a detachable or external USB drive for the server's own backup. I use a hot-swap drive in a 5.25" bay, which I leave connected most of the time, since backups are scheduled and automated.
But we all need to swap disks in and out.
What sort of procedure would guarantee this doesn't happen again?
I'd think one should probably remove the server backup configuration and disconnect the disk before adding new hardware, and then re-create the backup.
Certainly, someone who works in a data-center managing enterprise servers might have a clear idea of what is done routinely in those situations.
But I just had to post this thread with its questions. I might have completely missed the shortcut fix for it, and then reinstalled the OS and software from scratch.
Do people who work in data-centers managing servers ever thump their chests?