Just to start I really don't know much about Unix at all. I am the Novell and Windows administrator. I can create a printer and setup a backup job using ArcServe for Unix but that is really about it. OK about a week ago the backup job on our SCO Unix box was still running when I went to change to tape at 4pm (the backup was started at 11pm the night before). I had heard that if you get alot of cron jobs running that it will slow your server so we rebooted the server. In addition I changed the start time of our backups to 8pm. This morning I get in and am told be our users that the system is running really slow. I check the server and the backup is still running and it is almost 8am. Can anyone tell me where to start in troubleshooting this problem? I am completely in the dark here. Any help would be greatly appreciated.
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations gkittelson on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Slow Backup
- Thread starter kcookson
- Start date
In this case I would try a manual backup eg without using your program. Firstly you will need to stop your program attempting the backup for you...
You can use cpio or tar its up to you:
Login as root and type the following.
nohup tar cv8 / &
which runs the command tar cv8 in the background so you can leave it unattended....
If this works than I suggest you change your software. If it doesn't than we can try CPIO. If this doesn't work we are looking at changing tapes and drives.
If you havn't already done so I would also reboot the server, this resets the tape drive.
Andy
++
Hope this helps
You can use cpio or tar its up to you:
Login as root and type the following.
nohup tar cv8 / &
which runs the command tar cv8 in the background so you can leave it unattended....
If this works than I suggest you change your software. If it doesn't than we can try CPIO. If this doesn't work we are looking at changing tapes and drives.
If you havn't already done so I would also reboot the server, this resets the tape drive.
Andy
++
Hope this helps
Arc/serve is a good tool and is easy for operate
TAR is a easy command that only allows you to backup data files , with tar you CANNOT backup your whole system because tar is unable to backup empty files , directories and the /dev directory in other words just use tar for backup your non-root filesystems( /y , /u)
Command to make a backup with tar:
# tar -cv8 /u
To extract your tar backup use:
# tar -xv8
To see what is on the tape use:
# tar -tv8
Remember that with backups you have to take care with paths , i mean if your backup starts from the / that is a absolute path as my example above, but if you do not the / as your starting point for backup that is a relative path.
# tar -cv8 /x/data absolute path and when you extract this backup it does not matter where you are on the file system
#tar -cv8 x/data you must be at the directory taht contains the " x " subdirectory
the file that contains what is the relationship with the numbers of the tar caommand are found at /etc/default/tar file
CPIO is a another command for backup , with CPIO you can backup your whole system from the / .
To make a backup use:
#find /u -mount -depth -print | cpio -ocv -I/dev/rct0 for OpenServer
or /etc/rmd/ctape1 if you have UnixWare
to extract a backup use:
#cpio -icvdumB -I/dev/rct0 ...
To see what is on the tape use:
#cpio -itv -I/dev/rct0
The things regarding absolute or ralative paths is the same.
And just for the end the command dtype allows you to know in which format the archive was made.
#dtype /dev/rct0
I always know that the Server and all the hardware must be updated with the firmware and bios that helps with the performance.
I hope this can help
Champetu
TAR is a easy command that only allows you to backup data files , with tar you CANNOT backup your whole system because tar is unable to backup empty files , directories and the /dev directory in other words just use tar for backup your non-root filesystems( /y , /u)
Command to make a backup with tar:
# tar -cv8 /u
To extract your tar backup use:
# tar -xv8
To see what is on the tape use:
# tar -tv8
Remember that with backups you have to take care with paths , i mean if your backup starts from the / that is a absolute path as my example above, but if you do not the / as your starting point for backup that is a relative path.
# tar -cv8 /x/data absolute path and when you extract this backup it does not matter where you are on the file system
#tar -cv8 x/data you must be at the directory taht contains the " x " subdirectory
the file that contains what is the relationship with the numbers of the tar caommand are found at /etc/default/tar file
CPIO is a another command for backup , with CPIO you can backup your whole system from the / .
To make a backup use:
#find /u -mount -depth -print | cpio -ocv -I/dev/rct0 for OpenServer
or /etc/rmd/ctape1 if you have UnixWare
to extract a backup use:
#cpio -icvdumB -I/dev/rct0 ...
To see what is on the tape use:
#cpio -itv -I/dev/rct0
The things regarding absolute or ralative paths is the same.
And just for the end the command dtype allows you to know in which format the archive was made.
#dtype /dev/rct0
I always know that the Server and all the hardware must be updated with the firmware and bios that helps with the performance.
I hope this can help
Champetu
stanhubble
MIS
For troubleshooting:
if the job is still running try issuing the command
tape amount
this will report how much data has gone to the tape device and it doesnt matter if the job is still runnung.
check /usr/adm/syslog and /usr/adm/messages for any error messages....
are you backing up at lot more data all of a sudden?
check for disk usage....
and finally or something to try first
clean the tape drive.
HTH
stan
if the job is still running try issuing the command
tape amount
this will report how much data has gone to the tape device and it doesnt matter if the job is still runnung.
check /usr/adm/syslog and /usr/adm/messages for any error messages....
are you backing up at lot more data all of a sudden?
check for disk usage....
and finally or something to try first
clean the tape drive.
HTH
stan
- Thread starter
- #5
How much are you backing up? Is TD continually starting and stopping while you are backing up?
Sounds like you may have corrupted something , either in backup software, or in the tape controls.
Ed Fair
efair@atlnet.com
Any advice I give is my best judgement based on my interpretation of the facts you supply.
Help increase my knowledge by providing some feedback, good or bad, on any advice I have given.
Sounds like you may have corrupted something , either in backup software, or in the tape controls.
Ed Fair
efair@atlnet.com
Any advice I give is my best judgement based on my interpretation of the facts you supply.
Help increase my knowledge by providing some feedback, good or bad, on any advice I have given.
- Thread starter
- #7
We narrowed the problem down to the root directory taking excessive time (16 hours) to backup. We figured out that over the past couple of weeks the number of files in the root has gone from 47,700 to 287,000 files with the size increasing from 3GB to 4+GB. With pure dumb luck we found that var/opt/K/SCO/MMDF/2.43.36/spool/mmdf/lock/home/addr directory had 2,000+ msg.* files in it. We are removing the files but need to know why these files are being created. Also, we are running SCO ver. 5.0.5 on a Dell Poweredge 4300 if that helps.
Thank you very much for all of the help so far. It has really helped a lot.
Thank you very much for all of the help so far. It has really helped a lot.
stanhubble
MIS
its probably mmdf complaining that it can't send a cron produced message to user cron or sys or root, and then complaining that it can't send the complaint, and so on....
have fun
have fun
- Thread starter
- #9
- Status