Path: utzoo!attcan!uunet!cs.utexas.edu!rutgers!rochester!pt.cs.cmu.edu!rover.ri.cmu.edu!mkb
From: mkb@rover.ri.cmu.edu (Mike Blackwell)
Newsgroups: comp.sys.mac.programmer
Subject: Distributed data collection
Message-ID: <5167@pt.cs.cmu.edu>
Date: 8 Jun 89 23:20:46 GMT
Reply-To: mkb@rover.ri.cmu.edu (Mike Blackwell)
Organization: Carnegie-Mellon University, CS/RI
Lines: 22

I'm building a system which must collect data from a number of Macs, and
save it in a central database. The Macs are interconnected with AppleTalk.
The distrbuted Macs (perhaps 10 - 15 of them) only need to generate data (on
the order of 2 - 3 100 byte packets per minute -- slow), and the database
machine only needs to receive the data - information flow can be one
direction. The database file will need to be off-loaded daily, preferably
without having to shut the system down.

I've thought of two methods to acheive this. One is to have the database
machine simply be an AppleShare server. The clients then open, append, and
close the database file as needed. Simple, but I wonder how reliable this
will be - it seems that failure of one client can wedge the system, or
worse, corrupt the database. The second is to write a simple ATP
server/client setup where the server can receive packets and append them to
the database. This is more work to implement, but might be more reliable.

So my question - pros and cons to these two methods? Have I overlooked
anything? Your comments greatly appreciated.

		cheers, -m-

		Mike Blackwell		(mkb@rover.ri.cmu.edu)
--