Re: mksysb failure additional questions

From: Steve Thornhill (steve_thornhill@UK.IBM.COM)
Date: Tue Sep 24 2002 - 10:43:15 EDT


31776 of 37047 files (85%)..........................
0512-003 mksysb may not have been able to archive some files.

The 31776 is the number of files backed up at the 85% mark.

But print job files and stuff in /tmp are normal causes for a failure.

Steve Thornhill - IBM CLS
Int - 255104
Ext - 023 92 565104
Mob - 07778 289770
Email/Sametime - steve_thornhill@uk.ibm.com

                      Dave.
                      Zarnoch@sungard. To: aix-l@Princeton.EDU
                      com cc:
                      Sent by: IBM AIX Subject: mksysb failure additional questions
                      Discussion List
                      <aix-l@Princeton.
                      EDU>

                      24/09/2002 14:15
                      Please respond to
                      IBM AIX
                      Discussion List

Del and Folks,

Yes, I did check the archives and didn't see anything that was related.....

The way I see it:

There is a directory /opt/oracle that is 4GB

1 - If I exclude /opt/oracle, how confident are we that TSM backed up all
the files?

Here's an observation from an admin:

It just means that some files were open, and whats on the tape may not
match what was on the system. Unless you shutdown all applications, you
will get this message. Shouldn't be a message.

Could this mean that some Oracle files were open and not backed up by TSM
as well?

2 - Another observation:

Odds are it is files that were removed ( tmp files etc ) that were removed
between time the list of files to be backed up was created and the files
were actually being backed up.

This doesn't seem likely given the difference in files:

31776 of 37047 files (85%)..........................
0512-003 mksysb may not have been able to archive some files.
The messages displayed on the Standard Error contained additional
information.

5271 files removed between creating the list and the actual backup?

3 - Block count is incorrect...

Dunno, I'm set at 1024 which all my other mksysb are set at

4 - I think that we ran out of tape.....
     Unfortunately, the drive is a DDS2 which appears to only handle 4GB
     If this is true, and we decide to exclude /opt/oracle.....Refer to #1
above

Thanks!

Dave Zarnoch
UNIX Systems Engineer
SunGard eSourcing
600 Laurel Oak Rd.
Voorhees, NJ 08043
Dave.Zarnoch@sungard.com

Original email:

Folks,

We just started the initial mksysb of one of our servers

We are using a DDS2 120m tape drive and tape

Here's the filesystems in rootvg:

Filesystem 1024-blocks Used Available Capacity Mounted on
/dev/hd4 32768 9332 23436 29% /
/dev/hd2 786432 529908 256524 68% /usr
/dev/hd9var 458752 20772 437980 5% /var
/dev/hd3 131072 13384 117688 11% /tmp
/dev/hd1 32768 11916 20852 37% /home
/dev/lv00 1015808 623692 392116 62% /scs
/dev/lv01 14352384 3941976 10410408 28% /opt/oracle

Oracle processing is up.

Here's the end of the mksysb results:

21561 of 37047 files (58%)..............................
31776 of 37047 files (85%)..........................
0512-003 mksysb may not have been able to archive some files.
The messages displayed on the Standard Error contained additional
information.

I researched and found that a DDS2 tape should handle:

4GB with no compression

8GB with compression

According to SMIT, the tape drive is using compression

Can this be as a result of a bogus compression configuration or that the
Oracle
databases were up during the mksysb?



graycol.gif
ecblank.gif
pic18770.gif

This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 22:16:13 EDT