<p>What about the work directory?</p>
<div class="gmail_quote">On Apr 18, 2011 9:28 PM, "Mihael Hategan" <<a href="mailto:hategan@mcs.anl.gov">hategan@mcs.anl.gov</a>> wrote:<br type="attribution">> It should be a configurable parameter. I was thinking more like gpfs,<br>
> but it should be the user's choice.<br>> <br>> On Mon, 2011-04-18 at 18:53 -0700, Sarah Kenny wrote:<br>>> /var/tmp ?<br>>> <br>>> On Sun, Apr 17, 2011 at 1:25 AM, Mihael Hategan <<a href="mailto:hategan@mcs.anl.gov">hategan@mcs.anl.gov</a>><br>
>> wrote:<br>>>         I guess we should allow worker logs to go in some other place<br>>>         than<br>>>         ~.globus/<br>>>         <br>>>         Mihael<br>>>         <br>
>>         -------- Forwarded Message --------<br>>>         From: <a href="mailto:pads-users@ci.uchicago.edu">pads-users@ci.uchicago.edu</a><br>>>         Reply-to: <a href="mailto:pads-users@ci.uchicago.edu">pads-users@ci.uchicago.edu</a><br>
>>         To: <a href="mailto:pads-users@ci.uchicago.edu">pads-users@ci.uchicago.edu</a><br>>>         Subject: [pads-users] Mounting homes read-only on computes<br>>>         Date: Sat, 16 Apr 2011 22:00:40 -0500<br>
>>         <br>>>         -----BEGIN PGP SIGNED MESSAGE-----<br>>>         Hash: SHA1<br>>>         <br>>>         One of the things discussed as part of the post-mortem of<br>>>         today's home file server outage is mounting home directories<br>
>>         read-only on the compute nodes. We do this on Beagle and<br>>>         haven't had any complaints for the most part. So I wanted to<br>>>         solicit you to see how disruptive that might be to your<br>
>>         current PADS work. Homes would still be mounted read-write on<br>>>         the login machines, but they wouldn't be writable on the<br>>>         computes. The alternative is to push your writes to either<br>
>>         GPFS or the local scratch filesystems. If we don't hear any<br>>>         major complaints about this, we're targeting next month's<br>>>         maintenance to make this live. What are your thoughts on this?<br>
>>         -----BEGIN PGP SIGNATURE-----<br>>>         Version: GnuPG/MacGPG2 v2.0.14 (Darwin)<br>>>         <br>>>         iEYEARECAAYFAk2qV9gACgkQ4RgdOxQVi0CovACfW20YM0R8uexD5PeHoX6QGZc3<br>>>         LOwAnR3Iv1tokCpM81T56kSGrLNuniKg<br>
>>         =c5Y+<br>>>         -----END PGP SIGNATURE-----<br>>>         _______________________________________________<br>>>         pads-users mailing list<br>>>         <a href="mailto:pads-users@ci.uchicago.edu">pads-users@ci.uchicago.edu</a><br>
>>         <a href="http://mail.ci.uchicago.edu/mailman/listinfo/pads-users">http://mail.ci.uchicago.edu/mailman/listinfo/pads-users</a><br>>>         <br>>>         <br>>>         _______________________________________________<br>
>>         Swift-devel mailing list<br>>>         <a href="mailto:Swift-devel@ci.uchicago.edu">Swift-devel@ci.uchicago.edu</a><br>>>         <a href="http://mail.ci.uchicago.edu/mailman/listinfo/swift-devel">http://mail.ci.uchicago.edu/mailman/listinfo/swift-devel</a><br>
>> <br>> <br>> <br>> _______________________________________________<br>> Swift-devel mailing list<br>> <a href="mailto:Swift-devel@ci.uchicago.edu">Swift-devel@ci.uchicago.edu</a><br>> <a href="http://mail.ci.uchicago.edu/mailman/listinfo/swift-devel">http://mail.ci.uchicago.edu/mailman/listinfo/swift-devel</a><br>
</div>