krowitz@EDDIE.MIT.EDU@mit-kermit.UUCP (02/06/87)
There are a few more places were disk space can disappear to on
Apollo nodes. I have seen /sys/node_data/global_rws and
/sys/node_data/global_data grow to *very* large (3000 blocks or more)
sizes with no apparrant reason (no processes other than normal
servers running on the machine or its partners). If you are using
the VT100 emulator, you may find that your `node_data directory is
filling up with files like vte_ctl.?* and vte_mbx.?* These files are
small, but if it has been awhile since you cleaned them out there
can be a *lot* (hundreds!) of them. You can also try checking your
`node_data/proc_dir. I have occasionally seen this directory filled
with entries for which there was no corresponding process running on
the node.
-- David Krowitz
mit-erl!mit-kermit!krowitz@eddie.mit.edu
mit-erl!mit-kermit!krowitz@mit-eddie.arpa
david@mit-mc.arpa
(in order of decreasing preference)Erstad@HI-MULTICS.ARPA.UUCP (02/06/87)
What applications software is being run? This can also be a disk-space waster. For example, if you use Mentor tools I can list another bunch of places to watch out for.