Did you have any large log files that you have zeroed for space? If you have a process that is connected to a file, for example /var/log/apache/blah which was like 9370768, and you ran echo "" > /var/log/apache/blah
Even tho the filesystem shows 0 bytes or whatever, their is still open inode to that space on the filesystem, and it won't really be freed until no process is connected to the inode. You will have to HUP the process that is connected to the file, in this case apache. does that make sense? i know what i mean, but i dunno if i am explaining it right. -thor On Fri, 15 Mar 2002, John Hunter wrote: > > >>>>> "Brandon" == Brandon Harris <[EMAIL PROTECTED]> writes: > > Brandon> Linux sometimes has ghost processes that chew up large > Brandon> amounts of filespace and zombie out. Look for lots of > Brandon> zombie (Z) processes with "ps auxwwww" and see what that > Brandon> turns up. > > One: > > root 19153 0.0 0.0 0 0 ? Z 11:08 0:00 [httpd <defunct>] > > > But what would explain why I can't find the fucking files with du? > > I freed up some space by clearing out other files and the df numbers > went down commensurate with what I cleared. But the amount of free > space is steadily eroding. Something is eating it, but what and > where. May resort to the reboot after half an hour or so. Damned, my > cherished 122 day uptime is in jeopardy! > > JDH > > _______________________________________________ > Bits mailing list > [EMAIL PROTECTED] > http://www.sugoi.org/mailman/listinfo/bits > -- [[EMAIL PROTECTED] ]# _______________________________________________ Bits mailing list [EMAIL PROTECTED] http://www.sugoi.org/mailman/listinfo/bits
