does anyone here use scripts to monitor disk usage?
it would be nice to be alerted via email when the capacity of /dev/stripe/st0a goes above 90%.
also, it would be nice to alerted via email if gmirror or gstripe statuses change.
--
bsdnexus cross-post
it would be nice to be alerted via email when the capacity of /dev/stripe/st0a goes above 90%.
also, it would be nice to alerted via email if gmirror or gstripe statuses change.
--
bsdnexus cross-post
Something like
would work.
df | awk '/st0a/ { sub(/%/,"",$5); $5 >= 90 ? system("mail -s \"st0a < 90%\" me@mydomain.com > /dev/null"): 0 }'
would work.
and just run it daily via cron?
yep
how do you feel about this script?
#!/bin/sh
HOSTNAME=`hostname`;
HOME_USAGE=`df | awk '/usr\/home/ { sub(/%/,"",$5); print $5}'`;
GMIRROR_STATUS=`/sbin/gmirror status | grep "^mirror" | awk {'print $2'} | sort -u`;
if [ $HOME_USAGE -ge 90 ]; then
df -h | mail -s "$HOSTNAME: /usr/home usage high" root
fi
if [ "$GMIRROR_STATUS" != 'COMPLETE' ]; then
/sbin/gmirror status | mail -s "$HOSTNAME: raid array incomplete" root
fi
it seems to work well enough:
To: root@euterpe.wingedleopard.net
Subject: euterpe.wingedleopard.net: raid array incomplete
Message-Id: <20090816100000.982DD409D@euterpe.wingedleopard.net>
Date: Sun, 16 Aug 2009 04:00:00 -0600 (MDT)
From: lucas@euterpe.wingedleopard.net (▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇)
Name Status Components
mirror/gm0 COMPLETE ad4
ad6
mirror/gm1 DEGRADED ad8
ad10 (4%)
i deactivated ad10 to test the system.
now it's going to take an hour to sync up, which means i'll have thirty emails waiting for me in the morning (i'm running that script via cron every five minutes).
:)
now it's going to take an hour to sync up, which means i'll have thirty emails waiting for me in the morning (i'm running that script via cron every five minutes).
:)
i just got my first non-test (production) alert!
:)
time to buy some disks..
To: root@euterpe.wingedleopard.net From: lucas@euterpe.wingedleopard.net (▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇) Date: Sat, 12 Sep 2009 21:30:01 -0600 (MDT) Subject: euterpe.wingedleopard.net: /usr/home usage high Filesystem Size Used Avail Capacity Mounted on /dev/ad0s1a 496M 252M 204M 55% / devfs 1.0K 1.0K 0B 100% /dev /dev/ad0s1d 496M 32K 456M 0% /tmp /dev/ad0s1f 17G 3.1G 12G 20% /usr /dev/ad0s1e 496M 29M 427M 6% /var /dev/stripe/st0a 451G 371G 44G 90% /usr/home
:)
time to buy some disks..
the Rosewill RC-209-EX is $40 on newegg.com. :(
i should just invest in a new mobo (and cpu and ram and psu). sigh
i should just invest in a new mobo (and cpu and ram and psu). sigh
this is boring. atlas is too reliable. i never get emails from it. :(
atlas% crontab -l
*/05 * * * * ./bin/disk_alerts.sh
atlas% cat ./bin/disk_alerts.sh
#!/bin/sh
HOSTNAME=`hostname`;
HOME_USAGE=`df | awk '/usr\/home/ { sub(/%/,"",$5); print $5}'`;
GMIRROR_STATUS=`/sbin/gmirror status | grep "^mirror" | awk {'print $2'} | sort -u`;
if [ $HOME_USAGE -ge 90 ]; then
df -h | mail -s "$HOSTNAME: /usr/home usage high" root
fi
if [ "$GMIRROR_STATUS" != 'COMPLETE' ]; then
/sbin/gmirror status | mail -s "$HOSTNAME: raid array incomplete" root
fi
atlas%
test emails:
To: root@atlas.wingedleopard.net
Subject: atlas.wingedleopard.net: raid array incomplete
Date: Sun, 30 Jan 2011 01:34:00 -0700 (MST)
From: lucas@atlas.wingedleopard.net (▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇)
Name Status Components
mirror/gm9 COMPLETE ada0
ada1
mirror/gm7 COMPLETE ada2
ada3
mirror/gm5 COMPLETE ada4
ada5To: root@atlas.wingedleopard.net Subject: atlas.wingedleopard.net: /usr/home usage high Date: Sun, 30 Jan 2011 01:34:32 -0700 (MST) From: lucas@atlas.wingedleopard.net (▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇) Filesystem Size Used Avail Capacity Mounted on /dev/mirror/gm9s1a 496M 317M 139M 70% / devfs 1.0K 1.0K 0B 100% /dev /dev/mirror/gm9s1d 496M 22K 456M 0% /tmp /dev/mirror/gm9s1e 1.9G 13M 1.8G 1% /var /dev/mirror/gm9s1f 221G 1.6G 201G 1% /usr /dev/concat/gc9a 1.1T 948G 115G 89% /usr/home