AS/400 >> Backup - using SAVF & CPYSAVFDTA

by jchapman » Wed, 14 Jan 2004 11:51:19 GMT

Is anyone using SAVF as a method of backing up their iSeries?

If so, what are your experiences with it?

We are in the process of trying to reduce our downtime window. IBM has
suggested this method.

Any known problems with doing this?

Any Info would be greatly appreciated. Please email to my email shown
at the bottom.


Joe Chapman
Lead IT Specialist
Chatham Steel Corporation
Savannah, GA.

AS/400 >> Backup - using SAVF & CPYSAVFDTA

by sclind2 » Wed, 14 Jan 2004 12:32:00 GMT

What are you trying to backup - the entire system? That would take a lot of
savefile space.

How about using save-while-active?


AS/400 >> Backup - using SAVF & CPYSAVFDTA

by cle » Thu, 15 Jan 2004 05:59:08 GMT

I do.
Points to consider:
1- One savf can save 1 library only.
2- You need 2 jobs: one that creates savf + saves the library, the
other save the savf then destroy each of the savf when finished. The 2
jobs can communicate with each other using, let's say data queue.
3- Remember your first job may never fail, while the second can (tape
drive down for ex.). So the savf's can keep cumulating and eat up your
disk space. I have plenty of disk space (50%) , also have a the system
paging me when threshold is reached (90%).

AS/400 >> Backup - using SAVF & CPYSAVFDTA

by gb » Fri, 16 Jan 2004 01:48:51 GMT

We do!

We do it because we want to take the save-to-tape off the critical path
during our overnight processing rather than saving time.

We have a separate library that contains a*SAVF for each library that we
save. The backup job is two programs, one to save a list of libraries - one
to each *SAVF and the first job then submits a second job to save the *SAVFs
to tape.

This has several benefits:-

1) The time to backup the libraries is reduced (as the save to savf is
much quicker) so your post-save jobs (End Of Day etc) can be running
alongside the actual save to tape. This is what will save you the time and
is why IBM are recommending it.

2) The backup to *SAVF will not fail, the backup to tape may fail due to
media error(s). If, like us, you do not have operators on site during the
back-ups, your overnight processing will hang up waiting for the error to be
cleared. If you use the *SAVF route, the only job(s) to stop are those
requiring the tape drive. If we have media error(s) overnight, the operator
sorts out the problem with the tape/drive and then takes an R to continue
the save to tape. The save to tape takes negligible resource so it can
happily run alongside the normal daytime processing.

3) IF you have enough spare DSAD (as we do) then you can elect not to
clear some/all the save files so that you have an on-line backup of your
data - you don't even need the tape.


1) You must have enough disk space to hold the copy of the data i.e. the
*SAVFs. For those *SAVFs that we clear, we do it when the save to tape is
complete for all the libraries - it makes a restart easier if the media gets

Er, thats's it really...

Good luck!

AS/400 >> Backup - using SAVF & CPYSAVFDTA

by Steve Landess » Fri, 16 Jan 2004 13:14:15 GMT

If you follow this strategy, it will be faster and less likely to cause a
system failure if you install another disk controller and dedicate several
auxiliary storage units on this controller as another ASP and create a
library in this ASP for storing the save files.

Then, in the event you fill up this ASP, AFAIK it won't cause the entire
system to go down...



Similar Threads

1. Database backup without using 'backup database'

Suppose a running database (DB2 UDB) with all containers being SMS
based (file system as JFS or JFS2 on AIX).

How to make a backup (online) by doing a file system backup and not a
database backup?

Could this be done with:

db2 set write suspend for db

backup file system data used by db2 data and logs

db2 set write resume for db

with the purpose of doing:

restore of db2 file system data on another system
startup and usage of database.

The questions are:

    - will the restored database be consistent after start/restart

    - what additional resources are taken between suspend write and
resume write (a file system backup takes more time that just splitting
off a snapshot in a storage subsystem).  

Bernard Dhooghe

2. savact using go backup or ops nav? - AS/400

3. migrating from tape to backup-to-disk, using backupexec 12.5

Anyone familiar with Backup Exec?

I'm trying to wrap my head around how exactly I should be using 
'backup-to-disk' now that I've decided to ditch tapes here for my small 
office.  Running Backup Exec 12.5.

With tapes, we typically had 1 tape for each day of the week (plus weekends) 
and simply rotated through them.  We didn't pay much attention to overwrite 
protection time.  We ran full backups nightly, so each tape had a full 
backup from that day of the week.

I'm looking to move to disk, but am not quite sure how I should be setting 
it up.  Should I create one 'backup-to-disk' folder for each day of the 
week, and a different backup job for each day of the week?

Or should I have 2 folders, 1 for a full backup and another for 
incrementals?  And have 2 backup jobs, one that backs up full on Monday and 
the rest that backs up differentials every other day of the week?

Should I be using differential or incremental, what are the advantages of 
each?  Or should I be doing fulls for each day like I was with tape?

To create an off-site backup, can i just take a file copy of one of the 
backup-to-disk folders?

Basically I'm looking for some overall backup advice....

I'm backing up about 40GB of data per server (2 servers).

Thanks in advance.

4. Backup image file of Ghost 9 is bigger than used space on source disk C:\

5. db2 Backup using TSM

We are on Db2 v 8.2 for Linux

If we need to use TSM for backup of DB2, do we need to have TSM client
installed on the database server.

if yes, do we need to purcahse it separately or is it freely


6. Special considerations to backup using ntbackup

7. UDB: Backup Image Size & Used Pages Size

I posted this on other user-groups.
Hardly got any responses. Here goes the question:

PLATFORM = DB2 UDB 8.2+ on AIX 5.x, DPF

Two Topics broadly:
(A) I wish to find out a rough correlation between the size
     of Used Pages in the database and its backup image size
     (Have to do some space calculation).
(B) How many partitions should be backed up in parallel to
     gain the most optimum performance?

Specifically, the questions / observations on Topic A are
(Assuming Used Pages Size is fixed):
(1) Backup Image Size changes with number of buffers
     specified in the backup command.
(2) Is backup image size related to number of partitions
     even when the used pages are same?
(3) Does datatype have an impact on backup size?
If the used pages add upto 1G, approx how much would be the
size of uncompressed backup image?

All answers / hints / experience-sharing are appreciated.


8. Sample script for backups using DB2 v8 on Windows and TSM