Database Backup File much larger than database

[JSmith191]

system is Progress 9.1E running on a Windows 2003 Server.
About 6 months ago, our backup files on one database started hitting a size of over 20GB. We assumed this was due to the size of data in the database and started efforts to purge unnecessary data. 2 weeks ago, we did a dump and load in to a much smaller database (down to about 12GB of fixed space). The strange thing is that the backups are still of 21GB now. I'm stumped.

It is an online backup. Script and logs are below.
Has anyone else encountered anythin similar? Does anyone have any idea what I'm missing?

Script:
rem Syteline's DB Backup
@echo %Date% >> e:\backup\backup.log
call d:\DLC9\bin\probkup online e:\syteline.db\and01\and01 e:\backup\online\and01.bak >> e:\backup\backup.log
call d:\DLC9\bin\probkup online e:\syteline.db\supdb\supdb e:\backup\online\supdb.bak >> e:\backup\backup.log
call d:\DLC9\bin\probkup online e:\syteline.db\symglbl\symglbl e:\backup\online\symglbl.bak >> e:\backup\backup.log

From the database's log file:
23:00:01 BACKUP 28: Full backup started. (1362)
23:00:01 BACKUP 28:
1272145 active blocks out of 1500027 blocks in e:\syteline.db\and01\and01 will be dumped. (6686)
23:00:01 BACKUP 28: 4864 bi blocks will be dumped. (6688)
23:00:01 BACKUP 28: Backup requires an estimated 1.8 GBytes of media. (9285)
23:00:01 BACKUP 28: Restore would require an estimated 229705 db blocks using 9.7 GBytes of media. (9286)
23:00:01 BACKUP 28: Begin backup of Before Image file(s). (5459)
23:00:05 BACKUP 28: End backup of Before Image file(s). (5460)
23:00:05 BACKUP 28: Begin backup of Data file(s). (5461)
23:34:55 BACKUP 28: End backup of Data file(s). (5462)
23:34:55 BACKUP 28: Wrote a total of 37586 backup blocks using 9.7 GBytes of media. (9284)
23:34:55 BACKUP 28: Full backup successfully completed. (1364)

This is from the backup's log file:
Sun 06/03/2012
PROGRESS Version 9.1E as of Tue Oct 12 20:29:29 EDT 2004
1272145 active blocks out of 1500027 blocks in e:\syteline.db\and01\and01 will be dumped. (6686)
4864 bi blocks will be dumped. (6688)
Backup requires an estimated 1.8 GBytes of media. (9285)
Restore would require an estimated 229705 db blocks using 9.7 GBytes of media. (9286)
Backed up 1277009 db blocks in 00:34:54
Wrote a total of 37586 backup blocks using 9.7 GBytes of media. (9284)

 

[cj_brandt]

add the -com parameter to not backup blocks beyond the hwm and see if that makes a difference.

call d:\DLC9\bin\probkup online e:\syteline.db\and01\and01 -com e:\backup\online\and01.bak >> e:\backup\backup.log

 

[TomBascom]

How big is your bi file?

 

[Cringer]

How big is your bi file?

That was my thought too, but kind of discounted it as they've just dumped and loaded the db. It could have grown I suppose... so yes, worth an ask

 

[JSmith191]

The situation is resolved, but I'm at a loss to explain why. I deleted the .BAK files completely and waited for the normal batch backup routine to complete on its own. The files now reflect the amount of data that should be in the database. Bizarre, but I'm happy for the moment!