Xref: utzoo comp.sys.super:303 comp.unix.cray:284
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!sun-barr!lll-winken!llnl!kent@ocfmail.ocf.llnl.gov
From: kent@ocfmail.ocf.llnl.gov
Newsgroups: comp.sys.super,comp.unix.cray
Subject: file "families"
Keywords: Here at LLNL for years we have had the notion of file "families".  This
Message-ID: <800@llnl.LLNL.GOV>
Date: 21 Mar 91 22:14:54 GMT
Sender: news@llnl.LLNL.GOV
Reply-To: kent@ocfmail.ocf.llnl.gov
Followup-To: comp.sys.super
Organization: Lawrence Livermore Nat'l Lab.
Lines: 19

What makes a sequence of files a "family" is that at some levels the
sequence is considered to be one file.  The mechanism is that a sequence
number is appended to the name of the file, so the components might be
named file00, file01, file02, file03, and file0x, where the "0x"
indicates the last member.  We have library routines to facilitate the
manipulation of file families. 

I suppose you could think of file families as analogous to packetizing
data on a network.

Anyway, to the question:  has anyone on UNICOS (or other large machine
UNIX systems) done anything like this?  Has anyone else dealt with the
problems created by VERY large output files in some other way?

I would appreciate any info on this.  Please email to

kent@ocfmail.ocf.llnl.gov

If I get enough replies, I will summarize...