[PRad] prad raw data on tape

[Eugene Pasyuk]

Chao, 

Are you sending filtered files to the silo or keep them on disk? It makes sense to ask for new volume set on the silo and put files in different directory. Raw data volune set is b-prad-raw and directory /mss/hallb/praad/data/ 
For filtered data I suggest to use /mss/hallb/prad/pass0/ 

-Eugene 

> From: "Chao Peng" <pengc2010 at gmail.com>
> To: "Ashot Gasparian" <gasparan at jlab.org>
> Cc: "Eugene Pasyuk" <pasyuk at jlab.org>, "prad" <prad at jlab.org>
> Sent: Wednesday, June 22, 2016 2:14:44 PM
> Subject: Re: [PRad] prad raw data on tape

> Yes, I am replaying all the files, with 5 sigma level, size is reduced by a
> factor of 40.
> But the replaying process takes about 20 mins for 1 run, and getting file from
> silo takes much longer time, so it may need several days to finish.

> Best Regards,
> Chao

> On Wed, Jun 22, 2016 at 11:07 AM, < gasparan at jlab.org > wrote:

>> That is good,
>> Ashot

>> > We certainly will keep all tapes, there is no question about it. But once
>> > we have data rewritten with zero's suppressed we can move original tapes
>> > from the silo to the shelf and free up the space on the silo.

>> > -Eugene

>> > ----- Original Message -----
>> >> From: "Ashot Gasparian" < gasparan at jlab.org >
>> >> To: "Eugene Pasyuk" < pasyuk at jlab.org >
>> >> Cc: "prad" < prad at jlab.org >
>> >> Sent: Wednesday, June 22, 2016 11:00:56 AM
>> >> Subject: Re: [PRad] prad raw data on tape

>> >> Hi Eugene,

>> >> Yes, that is started by Chao. The question is, should we still keep
>> >> the data in original form? I would keep it if it is all possible.

>> >> Ashot


>> >>> I got some statistics on the raw data written to tape.
>> >>> There are 52 volumes in prad volume set. This is about 130 TB of raw
>> >>> data
>> >>> in 67200+ files.
>> >>> The first thing to do would be to run zero suppression on all files.
>> >>> This
>> >>> will make dataset mach smaller and easier to manage.

>> >>> -Eugene
>> >>> _______________________________________________
>> >>> PRad mailing list
>> >>> PRad at jlab.org
>> >>> https://mailman.jlab.org/mailman/listinfo/prad


>> _______________________________________________
>> PRad mailing list
>> PRad at jlab.org
>> https://mailman.jlab.org/mailman/listinfo/prad

> --
> Chao Peng
> Department of Physics
> BOX 90305
> Duke University
> Durham, NC 27708
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/prad/attachments/20160622/604548b3/attachment.html>