[d2n-analysis-talk] coda file copying policy on d2n

[Brad Sawatzky]

Hi Matt,

You may be watching this already, but please be careful that you do not
fill /data1 with copies from /cache.  We do not have enough disk on d2n
to hold all of our data files (something like 10TB, IIRC).

As a rule, we should be taking advantage of /cache whenever possible.
The distributed /cache FS has many times our paltry 2.4TB, and files
usually stick around for at least a week unless Hall B is going nuts
with replay run.

Copying files to /data1/raw_data/ makes sense for groups of files you
wish to run tests over some extended period, or if you are having files
expire from the cache before you can get to them.

Bottom line is, watch the disk space.  I'll be nuking files in raw_data/
if /data1 fills past 95%.

-- Brad

-- 
Brad Sawatzky, PhD <brads at jlab.org>  -<>-  Jefferson Lab / Hall C / C111
Ph: 757-269-5947  -<>-  Fax: 757-269-5235  -<>- Pager: brads-page at jlab.org
The most exciting phrase to hear in science, the one that heralds new
  discoveries, is not "Eureka!" but "That's funny..."   -- Isaac Asimov