BM> Is there a way (3.12) to set the server (32 meg, 4 gig x 2
 BM> mirrored) to automatically clean out old filenames in the
 BM> salvage system.  Out of all the servers I manage, there is
 BM> one that generates error
 BM> messages (sorry, can't quote it) after about 2 weeks, unless we
 BM> periodically run PURGE /ALL.
 FR> Volume ... is out of space?
No, lots of room on all volumes.
 FR> You could also set the retention time for deleted files to a smaller va
 FR> than the default.
Oooo.  I should have known this...where is it set?  A set xxxx
pararmeter?
 BM> I don't understand what algorithm Novell uses to manage the
 BM> memory-retained list, nor why this is the only server I've ever
 BM> encountered with this symptom.
 FR> Not sure exactly what you're referring to.  If it's that Netware doesn'
 FR> really delete file when you delete them, there are few others with this
 FR> feature.  When a volume gets low in space, the server will automaticall
 FR> purge files that are older than the retention time for the server.
Oh, miscommunication here.  I understand delete vs. purge, and
salvage.  I was saying I don't understand how Novell manages the
=LIST= of filenames that are deleted but still available for salvage. Where 
is that list kept (appears to be in memory)?  Other than a PURGE command, 
what causes names to be removed from the list, making that file no longer 
salvageable, and when/why does the removal of name(s) occur?
 FR> Deleted files are stored on disk.  [Amount of memory doesn't
 FR> determine the number] of deleted files that a volume can hold.
The miscommunication again.  Number of files (disk) that can be retained 
wasn't the issue; it is the number of fileNAMES (memory?).
Thanks for your input, Frank.  Changing the retention time to a lower value 
may solve this.
Bob
--- GEcho 1.11+
---------------