Is this how AI is supposed to work?

[ron]

Production: Sun V480, Solaris 8, Progress 9.1D06
Warm Spare: Sun E3500, Solaris 8, progress 9.1D06

We're testing a Warm Spare set-up using AI files to update the Warm Spare. There are 8 databases involved.

No problem on Production ... the AI files get swapped when they should ... and despatched to the Warm Spare server.

No problem in the Warm Spare applying the AI files -- works just fine.

No problem taking tape backups using probkup (with -norecover) on the Warm Spare server. And after taking a backup there is no problem continuing to apply the AI files.

HOWEVER ... if a backup from tape is restored to the Warm Spare it expects AI file #1 to be the next to be applied to it. That is not what was expected ... we thought we could continue to apply AI files in the same sequence as previously.

Is there any way to coerce Progress to NOT start back at #1?

If we have to have it the way it is -- it wrecks a substantial slice of our plans. (Where's the melting pot?)

Has anyone else found this to be a problem?

 

[ron]

I have been advised that the problem I have is "known" ... and the only solution is to take the backups on the warm spare server using an OS utility (eg, tar, cpio, ufsdump).

This is an extreme aggravation ... but if it has to be ... it has to be!

I should mention that taking a full backup using probkup (that, apparently, we will no longer be able to do) on a DLT70 tape drive took about 3.5 hours (120 GB). We have tried tar ... and it took over 24 hours! cpio is only a trifle faster! ufsdump may be our only practicable solution!! (We'll try it tomorrow.)

If ANYONE out there is taking backups off a warm spare system ... ** PLEASE ** contact me. I am very interested to compare notes. It seems that this is minefield territory!

 

[ron]

PROBLEM SOLVED

The problem we've hit is a bug introduced in 9.1D05 ... which is scheduled for fixing in SP07.

See P26847.