This is a script I use on all my machines. Feel free to hack as required. #!/bin/bash # the warning limit for disk usage in % DISK_USAGE=85 report() { echo $* # logger $* } # check disk usage full=0 for i in $(df | grep "/dev" | cut -c52-54) ; do if [ $i -ge $DISK_USAGE ] ; then full=1 fi done if [ $full -gt 0 ] ; then report "Warning: disk nearing capacity on `hostname`" df fi # check in use file handles count=$(cat /proc/sys/fs/file-nr | cut -f1) limit=$(cat /proc/sys/fs/file-nr | cut -f3) # do a simple check for 3 / 4 usage count2=$(expr ${count} \* 4) limit2=$(expr ${limit} \* 3) if [ ${count2} -ge ${limit2} ] ; then report "Warning: file handles nearing limit on `hostname` ( ${count} out of ${limit} )" fi # check allocates semaphore arrays eval `ipcs -ls | grep "max number of arrays" | tr -d " "` eval `ipcs -us | grep "used arrays" | tr -d " "` #echo $usedarrays #echo $maxnumberofarrays count=$usedarrays limit=$maxnumberofarrays # do a simple check for 3 / 4 usage count2=$(expr ${count} \* 4) limit2=$(expr ${limit} \* 3) if [ ${count2} -ge ${limit2} ] ; then report "Warning: semaphore arrays nearing limit on `hostname` ( ${count} out of ${limit} )" fi cat /proc/meminfo | grep ^Swap: | awk 'BEGIN {n = 3; d = 4} {total = $2 / 1024 ; used = $3 / 1024; if((total * n) < (used * d )) { printf "Swap usage is over %d/%d\n", n, d}}' Lamar Owen wrote: > On Friday 30 September 2005 04:21, Dag Wieers wrote: > >>On Thu, 29 Sep 2005, Lamar Owen wrote: >> >>>Oh, quick quiz: what happens to a CentOS box when you reboot after >>>running out of disk space on the / filesystem? You get a graphical login >>>just fine, but then you can't login (you login, and it returns you to a >>>login). Not hard to fix, but very mysterious to the newbie. > > >>I lost my Revelation password database because a disk-full condition >>recently. An 2 week old backup didn't have some important passwords. > > > Yeah, I lost some e-mail filters once that way, too, when my /home partition > filled up. A 'high-water' mark utility probably has been written, and I know > logwatch can do periodic e-mails that, if you take the time to read them, can > give you advance notice of a disk full impending, but there are those times > when the disk might get full quite suddenly (particularly the /var partition > if something creates log spew (which, with SELinux, can happen extremely > quickly)). > > >>BTW A tool that mails out when the disk is full is not a good solution... > > > No kidding. -- John Newbigin Computer Systems Officer Faculty of Information and Communication Technologies Swinburne University of Technology Melbourne, Australia http://www.ict.swin.edu.au/staff/jnewbigin