<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body text="#000000" bgcolor="#ffffff">
Hi Lamiaa,<br>
<br>
First your tcl file looks good.<br>
<br>
1. list of outputs are fine, but I think this time we should write
(i)<br>
and (ii) to tape silo. We may use them for quite some time. <br>
<br>
2. We will cook 10 files from each good run (1M or more triggers).<br>
But I want to ask to cook few 6 GeV runs in full. For example you<br>
can chose runs that were used for the calibration.<br>
Fill free to clean-up eg6 disk. We should remove all old data (bos,<br>
ntuple, histogram ...) files.<br>
<br>
3. Please use 64 bit far. We are done with 32 bit!<br>
<br>
4. Hovanes should be able to help on tagging the packages. I am<br>
working on some modification to seb for neutral trigeger. But may<br>
not be ready for this cooking. <br>
<br>
Regards, Stepan<br>
<br>
<br>
On 9/3/10 4:46 PM, Lamiaa El Fassi wrote:
<blockquote
cite="mid:AANLkTi=2F86LkNpXMWiEaZCLQ2NLOY-D33WJWwCQ33SS@mail.gmail.com"
type="cite">Hi Stepan and all,<br>
<br>
I would like to set up the strategy of the next pass0 cooking
based on what we have done earlier <br>
on the pass0_v0.<br>
<br>
1. List of output files that have been produced:<br>
<br>
(i) HBOOK files: SEB ntuples and ANAHIST files<br>
(ii) ROOT files : CLASTOOL & H2ROOT ntuples<br>
(iii) BOS files : Bosbanks format<br>
(iv) Helicity_correction files<br>
(v) LOG files : ASCII format<br>
<br>
In the previous cooking, <b>NONE</b> of these files was saved on
SILO. The files (i) and (iv) were saved on the permanent claseg6
work disk. However, all other files were stored on the
auto-manageable work <br>
disk. <br>
<br>
In this cooking, shall I produce the same output formats and save
them on the same place? <br>
<br>
2. Claseg6 work disk: the process of the same runs list (total of
796 runs) and ~ 10 files of <br>
each run (still to be confirmed) would amount approximately 1.8 TB
disk space on the permanent <br>
work area (based on 1). For the moment, 88% of a claseg6 disk
space is in use and it's partitioned <br>
as follow:<br>
<br>
(i) Previous pass0_v0: 1.8 TB <br>
(ii) DC_CALIB files : ~ 437 GB ! files used on the DC
calibration by Seema <br>
(iii) EC_CALIB files : ~ 387 GB ! files used on EC_time
calibration by Bayram<br>
(iv) ....Other cooking purpose covering... ~ 200 GB <br>
<br>
Before launching this new pass I need to clean some disk space, so
what shall I do with the <br>
previous cooking outputs? Please let me know if you have any
unused data to remove it.<br>
<br>
3. Which farm node shall we use on this cooking; 32 or 64 bit? I
was processing recently <br>
all the calibration jobs in the 64 bit farm nodes without any
problem (at least I didn't hear any <br>
complain from anyone!). <br>
<br>
4. I have updated the icf package with the new modification that
FX did by adding a correction <br>
on the vertex position. The new source code reside at
~claseg6/pass0_ect/. I will commit the <br>
new changes to the Eg6 CVS branch. The latter doesn't include all
the packages that have been <br>
modified like seb, icf & trk, is it possible to tag also those
packages on the Eg6 branch? <br>
<br>
Best regards,<br>
<br>
Lamiaa<br>
<br clear="all">
************************************************************<br>
* Lamiaa El Fassi email: <a moz-do-not-send="true"
href="mailto:elfassi@jlab.org">elfassi@jlab.org</a><br>
* Research Associate @ Rutgers University <br>
* Phone: (757) 269-7011 // Fax: (757) 269-5703 <br>
* Jefferson Lab., 12000 Jefferson Ave. <br>
* Suite# 4, MS 12H3<br>
* Newport News, VA. 23606<br>
************************************************************<br>
</blockquote>
</body>
</html>