Since each of the machines is handling only a part of the data, you
could have them use FTP to transfer individual files to a central
collection point, say the big mdisk currently in use. 

Regards, 
Richard Schuh 


-----Original Message-----
From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On
Behalf Of Ian S. Worthington
Sent: Wednesday, May 02, 2007 1:10 PM
To: IBMVM@LISTSERV.UARK.EDU
Subject: Re: Link Multiple Write safely

Hi Paul --

Many thanks for your considered note.

Our (big) shop does uses SFS, quite a lot.  I am in the process of
taking a
single process that has always used dedicated volumes and splitting it
to run
in multiple machines (as its now taking too long to finish (several
days) and
can easily be made to run in parallel).

There are clear and obvious benefits, as you and others have discussed,
to
putting these volumes into a private SFS pool but I've been told it
can't be
done.  The exact reasons have not been shared with me, but I do know a
brick
wall when I bang my head against one :)

So I'm looking for a different approach and a few of the ideas that have
been
passed around here may, with just a few changes, work well, and can be
done
without us having to request activities from IBM (and their beleaguered
support representative).

i

------ Original Message ------
Received: 
From: "Paul B. Nieman" <[EMAIL PROTECTED]>
To: IBMVM@LISTSERV.UARK.EDU
Subject: Re: Link Multiple Write safely

> Reading between the lines and drawing some conclusions...
> 
> You have several service machines that have some cylinders of
dedicated 
> minidisk space and write output files.
> From a later email, your shop uses SFS, but doesn't want you to use it
for 
> the LARGE amount of space that is currently dedicated to these service

> machines.
> > I'm told I can't have it for these (whole pack) volumes.  I'm not
sure of

> > the
> > reasons.  (WAG: Are there old technology disks that can't be
converted to
> > SFS?)
> 
> (The implication (to me) is that you could have it for smaller
minidisks, 
> but I may be reading too much into your words.)
> 
> Is it possible that your shop is falsely concerned about the amount of
space

> that will be needed for your SFS usage?  The minidisks used for the
output 
> files are likely over allocated to allow for size variations, so SFS
may 
> return some space for you (some of which you will have to use for the
SFS 
> control minidisks).  You should get a reason for your shop not
allowing the

> use of SFS.  Maybe you can allay their fears.  (No, to the WAG, AFAIK;

> someone else should answer that definitively.)  Naturally, there might
be a

> cutover period where a larger amount of disk space would be needed,
but you

> should be able to work through that without having to double the space

> allocated.  (Again, I'm assuming that there is a space concern.)
> 
> The use of the reader doesn't sound appropriate.  The large minidisks 
> implied suggests a high volume of write activity that would probably
not 
> perform well.  (At least, not as well as SFS.)
> 
> As others have said, SFS sounds appropriate.  If there are concerns
about 
> the performance of SFS, at least get that statement from your shop.
Without

> knowing why they are opposed to it, you can't build a justification.
> 
> While you are at it, some clarification of the current activity would
also 
> help, to better understand the problem.  Actually, why isn't the
current 
> state (writing to separate minidisks) no longer acceptable?  What are
they 
> trying to achieve by combining them (space savings ;-) ?  Or improved 
> cleanup (another reason for SFS, IMO)?  Easier browsing (put the files
into

> directories named after the originating service machines and it is
easy to 
> browse with DIRLIST)?  This last (separate directories) also
alleviates any

> duplicate filename conflict problems that other solutions might have.
> 
> 
> ----- Original Message ----- 
> From: "Ian S. Worthington" <[EMAIL PROTECTED]>
> To: <IBMVM@LISTSERV.UARK.EDU>
> Sent: Tuesday, May 01, 2007 1:00 PM
> Subject: Re: Link Multiple Write safely
> 
> 
> That's interesting, but conflicts with what some other folks have said
here
> earlier.
> 
> We are under CMS, and I don't want to do simultaneous update to a
single 
> file,
> just allow several service machine to write their output files to
single
> location, which requires a mw-tolerant directory.
> 
> Apparently we can't convert the minidisks we're using to SFS.
> 
> NFS, as mentioned earlier might be an option: I'm trying to find out
if its
> enabled.
> 
> ian
> 
> ------ Original Message ------
> Received: Tue, 01 May 2007 05:29:09 PM BST
> From: "Schuh, Richard" <[EMAIL PROTECTED]>
> To: IBMVM@LISTSERV.UARK.EDU
> Subject: Re: Link Multiple Write safely
> 
> > There is a very big one at this shop that is mode 6, and has been
shared
> > for the 9+ years that I have been here and, I think, about 6 years
> > before that. With mode 6, you can update in place so long as you do
not
> > change the record length of an existing record. We have a source
control
> > system that uses ISPF panels that requires that all users get the
disk
> > having this pseudo maclib MW.
> >
> > In any event, the updating of a record in place requires no changes
that
> > will corrupt the disk if it is MW by several users, in our case,
over
> > 50, at the same time. So if that is how you intend to use the disk,
MW
> > is perfectly safe under CMS.
> >
> > Regards,
> > Richard Schuh
> >
> >
> > -----Original Message-----
> > From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED]
On
> > Behalf Of Ed Zell
> > Sent: Tuesday, May 01, 2007 9:06 AM
> > To: IBMVM@LISTSERV.UARK.EDU
> > Subject: Re: Link Multiple Write safely
> >
> > > Mode 6 files can be shared so long as the software takes
> > > care not to have record collisions. ISPF handles sharing
> > > its fake MACLIBs that way under CMS.
> >
> >
> > I don't think this is the case.  None of our ISPF MACLIB's
> > are file mode 6.
> >
> > I was under the impression that ISPF handles sharing by forcing
> > all updates to go through the ISPVM virtual machine.  But I
> > could be wrong (as has happened many times in the past!).
> >
> >
> > Ed Zell
> > Illinois Mutual Life
> > (309) 674-8255 x-107
> > .
> >
> >
> > CONFIDENTIAL NOTICE:  This communication, including any attachments,
is
> > intended only for the use of the individual or entity to which it is
> > addressed and contains information which may be confidential.  If
you
> > are not the intended recipient, any distribution or copying of this
> > communication is strictly prohibited.  If you have received this
> > communication in error, notify the sender immediately, delete the
> > communication and destroy all copies. Thank you for your compliance.
> > = 
> 

Reply via email to