topic Re: success of fbackups in Operating System - HP-UX

success of fbackups

SUBRAMANYAM ANUPINDI_1 — Wed, 31 Mar 2004 15:38:21 GMT

Hi,

I st server backup over network
-------------------------------
I am taking backup of all mount points including vg00 of our HP server using fbackup over network with following command :

/usr/sbin/fbackup -c /var/adm/fbackupfiles/config.backup -g /var/adm/fbackupfiles/PSDEV_full -0v -u -f psdw:/dev/rmt/0m -I /backup/psdev_full.index

The config.backup contains following info :

blocksperrecord 256
records 32
checkpointfreq 1024
readerprocesses 6
maxretries 5
retrylimit 5000000
maxvoluses 200
filesperfsm 2000

After backup the output in "dates" file shows as:

/var/adm/fbackupfiles/PSDEV_full 0 1080593253 1080616731
STARTED: Mon Mar 29 16:47:33 2004 ENDED: Mon Mar 29 23:18:51 2004

After backup following errors occured :

fbackup(3050):total file blocks read for backup:107581488

fbackup(3056):total blocks written to outputfile psdw:/dev/rmt/0m:0(zero)

fbackup(1030):warning encountered during backup

Questions:
----------
1) Why total blocks written to tape(psdw:/dev/rmt/0m) shows up as 0(zero)

2) what does two values "1080593253" "1080616731" signify from dates file.

3) I want to know if my backup is good enough for recovery.

II nd server backup with local tape drive
-----------------------------------------
In another backup of another server which has local tape drive i got following error message:

fbackup(1102):WARNING: unable to stat file /var/spool/sockest/pwgr/client 3314

fbackup(3005):WARNING: file number 157737 was NOT backed up

question
-------
1)How to identify file number 157737 ?

2) What these two errors mean to integrity of backup

3) Can i recover files backed up after error fbackup(1102) and fbackup(3005).

4)I want to know if my backup is good enough for recovery

Thanks

S.Anupindi

Re: success of fbackups

Sundar_7 — Wed, 31 Mar 2004 17:38:17 GMT

Hi Anupindi,

I can try answering couple of your questions.. :-)

1)what does two values "1080593253" "1080616731" signify from dates file.

The values represent the date in UTC format. (i.e from Jan'1 1970)

# echo "0d1080593253=Y" | adb

use the above command to make sense out of it

2) I want to know if my backup is good enough for recovery

Try using frecover and list the files in the tape from the remote system

3)fbackup(1102):WARNING: unable to stat file /var/spool/sockest/pwgr/client 3314

fbackup(3005):WARNING: file number 157737 was NOT backed up

/var/spool/socket/pwgr/client is a socket file created by the password caching daemon. fbackup creates the list of file to backup at the beginning. So what has happened is that, when fbackup was started, it saw this socket file but when it tried to backup, this socket was already closed/removed.

So you can safely ignore this error message.

Sundar

Re: success of fbackups

SUBRAMANYAM ANUPINDI_1 — Thu, 01 Apr 2004 20:47:59 GMT

Hi All,

Sundar,Thanks for posting reply to my message.

I checked with frecover -I option to a temp file and see the files. In fact i tried by deleting a small folder and restored it over network.

I am not clear of two error messages :

In first backup :

fbackup(3056):total blocks written to outputfile psdw:/dev/rmt/0m:0(zero)

I am not clear why it shows zero for output file ?

In second backup How to identify file number 157737 ?

Thanks

S.Anupindi

Re: success of fbackups

SANTOSH S. MHASKAR — Fri, 02 Apr 2004 02:18:52 GMT

Hi

Answer for ur question

How to identify file number 157737 ?

When backup proceeds fbackup gives number to each file it is backing up. 157737 is the file no. Just go to Line 157737 in ur Index file
, this is the file. Which is not backud up hence ur backup is not complete & this file cannot be recovered.

Also when fbackup gathers info about file to backup the file named

/var/spool/sockest/pwgr/client

existed but at the time of backup it was not there hence the error message this file also u cannot recover.

-Santosh

Re: success of fbackups

john korterman — Fri, 02 Apr 2004 03:44:26 GMT

Hi,
regarding the number of bytes written...
It has to do with the fact that the backup is written to a device on a remote machine.
Please check this link:
http://www4.itrc.hp.com/service/cki/docDisplay.do?docLocale=en_US&docId=200000072010909

If the link is not available, then go to the technical knowledge base and search for Document id: 8606344655

It is apparently a known problem, but so new that there is not yet any solution attached. However, it may come.

regards,
John K.