User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8.1.12) Gecko/20080201 SeaMonkey/1.1.8
I forgot to bring this topic up during today's EG1 meeting, but also want everyone who was absent to have a chance to respond.
Basically, I think we could offer them the following:
1) Remove all EG1a and EG1b raw data tapes from the silo whenever they want to -> we will not recook (the tapes would be put in storage, so in a real emergency they could potentially still be put back in service)
2) Keep the cooked BOS files for at least one more year, in case we need to recreate DSTs from them
3) Keep ntuple10s for another 2 years - they are still useful for us as the recent "wiggle study" just showed
4) Keep DSTs for at least 5 years and then revisit this issue. They should not be removed until there is no chance that another analysis might use them - we still have a lot of ongoing analyses and plans for future ones.
Please give me your feedback on this proposal - I should respond "soon". - Sebastian
Hi Sebastian,
The issue of computing resources by eg1 came back.
Any new thoughts on what you want to do.
Thanks, stepan
-------- Original Message --------
Subject: [Fwd: Re: CLASEG1 computing and storage requirements?]
Date: Wed, 20 Feb 2008 15:23:31 -0500
From: Sandy Philpott <sandy.philpott@jlab.org>
Reply-To: Sandy.Philpott@jlab.org
To: Stepan Stepanyan <stepanya@jlab.org>
Also we'd like to revisit the eg1 requirements as noted from this last
discussion...
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.4) Gecko/20011126 Netscape6/6.2.1
Sandy,
We discussed tape and disk storage issue on today's eg1 meeting.
Group wants to leave every thing as they are. There have been
some new developments in the event reconstruction and eg1 may
re-cook some their data. In general group is quite active in physics
analysis and although they recognized that some of the files were not
accessed for long time, they want to have them accessible as they
are.
So, I think what we want to ask is to leave things unchanged and may
be come back to this discussion in 1/2 year or so.
Regards, Stepan
Sandy Philpott wrote:
Hi Stepan,
Our records show that you own the user account for CLASEG1. We are gathering
computing and storage resource requirements from the experiments for the
nearterm and longterm, and would like to get your feedback on EG1's future
needs. (Are you the right person for this?)
Currently EG1 has used about 1000 tapes in the mass storage system, 2.25TB
of cache disk, and less than 1% of the farm compute capacity since we began
tracking by project in midsummer. I'm not sure how much work disk EG1 uses
- may have to check with Mark Ito for that information. I have included information
we have about EG1 usage at the bottom of this message.
What are EG1's plans for the future? Can we remove these tapes from the
mass storage system and keep them in our tape vault, or are they needed online?
Thank you,
Sandy Philpott
---------------
Family claseg1 Cache Usage Summary:
Usage : 74.89% of 3000000000000 alloted bytes in use
Pins : 0.00% of allotted bytes are pinned
Reserved : 0.00% of allotted bytes are reserved
Median Age : 56d 23:28
10% Age : 4d 21:29
90% Age : 455d 0:11
