hbrady
asked on
Corrupt Q_xxx.srv files
We have a Novell 4.11 server running on a Micron NetframeMV5000. The server also has sp6 installed. I'm am not at the office so my exact figures may be off; however, I will try to be as accurate as possible.
Approx. 3 months ago a problem started that has repeated itself every 19-30 days. Our Novell server will crash with a bunch of FAT errors on the console screen. A reboot shows several "truncated" file errors (scroll by so fast that I can't read them) ; however, a check of the log files shows them to be Q_xxxx.SRV files. These are the same files mentioned in the volume error logs at the time of the crash.
After a reboot the server appears to be OK for several minutes; however, the FAT errors start again and all clients loss their connections. If I do a purge /all on SYS the server will stay up and everything looks fine the next reboot will also look clean.
I'm confused. I have no idea what a Q_xxxx.srv file is or why it would cause my Novell server to occasionally have a brain seizure. Anyone have any ideas ? I am a Linux and NT person who is NetWare Challenged. :-)
Approx. 3 months ago a problem started that has repeated itself every 19-30 days. Our Novell server will crash with a bunch of FAT errors on the console screen. A reboot shows several "truncated" file errors (scroll by so fast that I can't read them) ; however, a check of the log files shows them to be Q_xxxx.SRV files. These are the same files mentioned in the volume error logs at the time of the crash.
After a reboot the server appears to be OK for several minutes; however, the FAT errors start again and all clients loss their connections. If I do a purge /all on SYS the server will stay up and everything looks fine the next reboot will also look clean.
I'm confused. I have no idea what a Q_xxxx.srv file is or why it would cause my Novell server to occasionally have a brain seizure. Anyone have any ideas ? I am a Linux and NT person who is NetWare Challenged. :-)
Sounds like you have a drive error or controler failure.
Usually when a drive fails, queue files are the most likely to be open at any time. As a part of the automatic checking on restart, they will always report an error eg truncated etc. This is not part of your problem - just a symtom of the server not shutting down correctly. The same could be said for the errors when the server starts to crash.
Go to the monitor screen on the server - if monitor is not loaded - load by typing load monitor at the : prompt screen
Move down to Disk information
Select the drive
Once at the drive screen, push tab to display the whole window. You should see and entry for redirected blocks and redirection blocks.( assuming you dont have a raid array)
Redirection blocks should be about 1200 or so ( depends on drive size)
Redirected block should be small number if any.
Note the number of redirected blocks and recheck next time the server fails. If the number goes up, suggests hard drive failure.
Note. Novell reserves a number of blocks on the drive to be used if a block can not be read. eg from bad hardware,disk etc. If it finds a bad block, it redirects the system to a block in the redirection blocks.
Usually when a drive fails, queue files are the most likely to be open at any time. As a part of the automatic checking on restart, they will always report an error eg truncated etc. This is not part of your problem - just a symtom of the server not shutting down correctly. The same could be said for the errors when the server starts to crash.
Go to the monitor screen on the server - if monitor is not loaded - load by typing load monitor at the : prompt screen
Move down to Disk information
Select the drive
Once at the drive screen, push tab to display the whole window. You should see and entry for redirected blocks and redirection blocks.( assuming you dont have a raid array)
Redirection blocks should be about 1200 or so ( depends on drive size)
Redirected block should be small number if any.
Note the number of redirected blocks and recheck next time the server fails. If the number goes up, suggests hard drive failure.
Note. Novell reserves a number of blocks on the drive to be used if a block can not be read. eg from bad hardware,disk etc. If it finds a bad block, it redirects the system to a block in the redirection blocks.
Another suggestion is set all your spool directories as purge imediate so it doesn't keep the spool file there after the deletion. What type of drives are they? Is this a server with server class drives or a workstation made into a server? How much free space is on the drive in question and how many free directory entries do you have. I would be getting replacement drives and preparing for the replacement.
ASKER
Thanks everyone. This a project that can only be worked on during OFF hours so I will let you know what happens when I get a chance to work on it.
Redirection blocks is 1147
Redirected blocks is 0
Thanks again
Redirection blocks is 1147
Redirected blocks is 0
Thanks again
Try the following
1) unplug and plug SCSI cables
2) Check termination
3) If you use Adaptec card add tag_disable=ffff when loading drivers option
4) Update disk drivers
5) swap out scsi card
1) unplug and plug SCSI cables
2) Check termination
3) If you use Adaptec card add tag_disable=ffff when loading drivers option
4) Update disk drivers
5) swap out scsi card
jstegall has the right answer. Seting all the queue directories to purge immediate will solve the problem. We had a similiar proglem using Netware 4.1.
If jstegall has the right answer you are running out of disk space or directory entries and the disk/fat is full before the files that have been printed become purgable. Setting this can only be a short term fix. You either need a bigger disk or you need to clean up your disk. To check the directory entries right click on the volume object in NW admin, select detail and then statistics look at the directory entries, if that is above about 90% thats quite bad.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Try to dismount the volume and run vrepair.
At the server console type "dismount <volumename>" without "" and then "load vrepair" also without "".
If the volume is sys and you have namespaces loaded you ned to load the namespacesupport files and vrepair before you dismount the sys volume (for example v_long.nlm or v_mac.nlm if yoy have long or mac support on the volume)