It's possible that HPFS is hanging ADSM.
One suggestion I have is to, if possible, re-format the
partition with the FORMAT command and /L parameter.
This will note bad spots on the disk. It's possbile that
the corruption is due to bad spots not detected when the
partition was first formatted because it does a "quick" format
by default.  And when system files end up on bad spots,
all kinds of strange things can happen.

Barry Fruchtman
ADSM Development




On Fri, 17 May 1996, Bryan King wrote:

> On May 3 I wrote that I was experiencing OS/2 server lock ups
> during an ADSM backup...
>
> This past weekend we discovered that there was actually
> corrupted entries/files on the drive - FDISK and a restore
> from ADSM solved this problem - minus two days of data.
>
> My BIG problem now is that another server is experiencing the
> same type of problem.  During the night backup, ADSM happens
> onto a what I suspect is the corrupted file, and the console
> displays an "HPFS error 0101" and practically halts.  Users are
> unable to login, and it is dead.
>
> Some staff would like to point fingers at ADSM, but I keep
> insisting that if the file system was not corrupt, ADSM would
> work fine.  Fix the file system, and we would all sleep at
> night.
>
> Their argument is that other applications will terminate
> without hanging the server, so ADSM must have a bug.
>
> The million $$ question is:
>
> Why doesn't ADSM skip over corrupt files, as if it is open?
>
> Is there a patch for this?
>
> The OS/2 client is running the March 26 release client.
>