Ok, few things to check... misc priorities of different tasks... even
     though I believe that a recall/recover will,  by default have a higher
     priority than a migrate so it should terminate the migrate and load
     the necessary tape for the recall.... also look into the mount limit
     for your 3590 device... if you have multiple drives, set the mount
     limit equal to the number of drives... ie in our atl we have two
     drives and the mount limit on our 3590dev is 2.  Now, correct me if
     I'm wrong but wouldn't the proper sequence be backup/archive to the
     disk pool and then migrate it off to the tape pool... and then based
     on the amount of data being pushed onto the server and the size of the
     disk pool you need to have your diskpool high migration value set to
     where it will start to bleed the data off early enough to prevent the
     disk pool from becomming 100%full, if this happens your backups will
     switch to the NEXTPOOL to write and I believe the backup/archive
     process will have a higher priority than the restore/recover and you
     would have to wait on backups to finish.... example if you have a 10GB
     disk pool and you have 5 processes archiving @ 1 GB /hr then the disk
     pool would be filling at 5GB/hr then in 2 hrs your pool would fill...
     if yo started bleeding the data off at 2GB/hr once at 80% full it
     would only buy you less than 1/2 hour so in 2.5 hours you would be
     writing backup/archive data straight to tape... if this is the case
     you would need a bigger disk pool...
     BIG OPTION would be to turn on cache for your disk pool, then even
     when the data is transfered to tape the copy on disk would remain
     active/available until it is necessary to overwrite it with additional
     backups..... Check into a few of these options and see if any would
     help...
     later
            Dwight



______________________________ Reply Separator _________________________________
Subject: Re: ADSM Error Message ANS4312E
Author:  ADSM-L at unix,mime/DD.RFC-822=ADSM-L AT VM.MARIST DOT EDU
Date:    6/27/96 7:07 AM


Hi,
        Thank for your reply. As we investigate further, I think our problem is
related to how the files are being
migrated from DISTRIBUTION-DISK to DISTRIBUTION-3590, as the file being archive
onto DISTRIBUTION-DISK as it's being restore on to disk by the application progr
am (some tape to disk read/write program, I'm not sure exactly, most likely SEGY
 read).
      As the file is being put on system disk
it's then being archive (dsmc archive ...) and set access (dsmc ...), and these
will be stored in DISTRIBUTION-DISK, but as the disk overflowed, these files is
being migrated to DISTRIBUTION-3590
Now this is when we have problem with a particular file.
     I think the problem could be that the file is being
migrated to tape somehow could not being migrated. The perculiar thing is this i
s the only file in the
sequence that have problem with, files before or after did not have any problem.

        Dwight, you may be right that we have problem with the mount wait time a
nd so on. Could you or any one show me how to circumvent this problem in the fut
ure.
      Thanks a lot. Paul.