NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
FYI: regarding the scour script.On linux a simple modification to the scour script can allow you to remove files on the order of minutes (instead of the default days). It also works on Solaris if you install gfind and use it as opposed to the Solaris find command.
The change is:
original:
find . -type f -mtime +$FINDAGE -name "$pattern" -exec rm -f {} \;) \
new:
find . -type f -mmin +$FINDAGE -name "$pattern" -exec rm -f {} \;) \
The just modify your scour settings to be minutes not days.
This has worked very well on many of my systems where I ony keep 1-4
hours of data at a time.
Tom Yoksas wrote:
From: Robert Mullenax <rmullenax@xxxxxxxxxxxx> Organization: UCAR/Unidata Keywords: 200405051451.i45Ep0tK004901 IDD CRAFT hardwareHi Robert,Can some of you that are receiving CRAFT data and using dcnexr2 to decode it give me an idea of what hardware setup you have (CPUS, disk, RAM) and how many sites you are receiving.The UPC currently ingests and decodes (uncompresses and files via dcnexr2) all CRAFT sites on two different PC platforms: FreeBSD (4.9-STABLE) 2 x Athlon 2400+ Tyan SN2466N-4M motherboard 2 x 1 GB RAM 300 GB hardware RAID (Adaptec RAID card) Fedora Core 1 (2.4.22-1.2188.nptlsmp) 2 x Athlon 2800+ Tyan SN2466N-4M motherboard 3 x 1 GB RAM (was running with 1 GB until Monday) 500 GB software RAID (my experience with the Promise Tech TX2000 under Linux was not rewarding) Both systems are ingesting ALL datastreams available through the IDD and decoding everything with GEMPAK decoders, most with McIDAS-XCD decoders, several with our netCDF decoders, and running the ADDE remote server. Our experience is that the FreeBSD platform performs measurably better than the one running Fedora Core 1. This may or may not have something to do with the hardware RAID on the FreeBSD system. I have been experimenting with various tunings on the Linux system to see if we can better performance, the most notable of which was writing a scouring script in Tcl and using it to prune the number of files kept in the nexrad/craft (64), nexrad/NIDS (64), and images/sat (96) directory trees used by GEMPAK decoders. Cheers, Tom -- +-----------------------------------------------------------------------------+ * Tom Yoksas UCAR Unidata Program * * (303) 497-8642 (last resort) P.O. Box 3000 * * yoksas@xxxxxxxxxxxxxxxx Boulder, CO 80307 * * Unidata WWW Service http://www.unidata.ucar.edu/* +-----------------------------------------------------------------------------+
-- Karen.Cooper@xxxxxxxx Phone:405-366-0434 Cell:405-834-8559 SAIC/Systems Analyst National Severe Storms Laboratory
ldm-users archives: