SPELL=procstatd VERSION=1.4.2 SOURCE=$SPELL-$VERSION.tgz SOURCE_HASH=sha512:e6ae2d22a04b966e2eff3044196531d772bb569798ea5ef61644d2a5a0044f8bad09ad8e24eb77328a0782dbae5599979b3ca817cb2b349469077fbc21394dca SOURCE_DIRECTORY=$BUILD_DIRECTORY/$SPELL-$VERSION SOURCE_URL[0]=http://www.phy.duke.edu/brahma/$SOURCE LICENSE[0]=GPL KEYWORDS="cluster" WEB_SITE=http://www.phy.duke.edu/brahma ENTERED=20020609 SHORT="daemon for gathering statistics about /proc" cat << EOF procstatd is a daemon that can be run either from userspace or inetd that can be queried via a simple API to parse various files in /proc and transform them into statistics. The returned data is in a simply parsable ASCII string (in a single packet) and might be used to, for example, drive a simple display. The nice thing about it is that by running procstatd's on all nodes of a beowulf or cluster, users or managers can (with suitable front-end tools) see what the loads and so forth are. The package contains also a perl script (webwatchd) for generating a simple html page with the statistics of the nodes. EOF