Emergency UNIX and Linux Support

Sorry, I though it was apropos the original chain: Just a note about best practices to install once the author hits a lull in his storm. Of course, if he needs help writing them, each is certainly another chain.

I guess if you have a file open for write under a mount, it can be mysterious when the space is used. Can mount point dirs be 555 to prevent entries?

Sometimes deleted open for write can also 'hide' a growing space hog, depending on how the FS handles that (*/lost+found/ or generated file name or ???).

Swap can make the device shrink, something we often fought with on Solaris with /tmp.

BTW, I have a mixed mode (files and subtrees) script call 'biggies' to locate problem areas (if there is enough disk somewhere to sort) that does something like this:

Code:

(
du -k "$@"
find "$@" -type f |xargs -n999 du -k
)| sort -nr | more

@DGPickett
The original problem was about a full filesystem. Please read up about the "-xdev" parameter to "find", you may find it useful.

Please check for any hung process.

One similar issue occured with us.
Our SFV-490 server "/ file system" got full.
We checked for /var directory and any other large file, but we could't find anything.

Only /proc directory was showing 1.5T in size.we checked that directory and found the size of the directory of PID of "fmd" process was having the size 58 GB. Then we just restarted the "fmd" process and then our / file system was back to normal condition realeasing a total space of 40 GB.


Cheers.