henry@utzoo.UUCP (Henry Spencer) (08/02/83)
/dev/usr ran out of space early this afternoon. Many thanks to Laura
for spotting it quickly and for her help in containing the disaster.
The only really satisfactory permanent solution for it is more disk,
i.e. a two-drive system, but as a temporary measure I have done
a quick kludge: /dev/usr has been enlarged by absorbing /dev/broot.
We are now running a new /unix with the appropriate changes to the
configuration tables. Fortunately /dev/broot immediately followed
/dev/usr, so I could simply patch the size number in usr's superblock
and have fsck pick up the newly-added blocks. This is not quite as
good as a mkfs followed by a full restor, because the filesystem does
not have as many inodes as it should have for its size, but that is
not a serious problem because the V7 mkfs allocates rather more inodes
than are needed anyway. (The inode table on usr is only about half
full despite zillions of small news-article files.) And patching the
superblock is several orders of magnitude faster than a full restor...
The loss of broot is unlikely to be a problem. It got very little use
anyway, and most of what it did get was just a matter of convenience.
It's unfortunate that there is no similar idle area immediately above
/dev/zoo, which could use some enlarging.
The schedule for a two-drive system is still unsettled because of
persistent hardware problems.
--
Henry Spencer @ U of Toronto Zoology
{allegra,ihnp4,linus,decvax}!utzoo!henry