Scripts for collectd : HP Proliant G6, ping, ZFS
Uses hpasmcli to read : Fans, Power usage (power meter in PSU), Temperatures (22 readings)
To monitor the array, but work in progress (stand-still).
You can ping a few hosts with this script. The build-in ping doesn't match the system pings.
This is also in progress but can report :
in bytes :
pool size
pool allocated
pool free space
pool scrubed
pool scrub repaired
in % :
pool usage
pool unused
pool scrubed %
in boolean (1=ok, 2=BAD)
pool health
in bytes/sec:
pool scrub speed
in time:
pool scrubed total time
pool scrub remaining time
in int:
pool scrub repaired errors
Usage : In /etc/collectd/collectd.conf:
<Plugin exec>
Exec "daemon" "/usr/local/bin/ping.sh" "8.8.8.8" "google.com" "opendns.com" "80.80.80.80"
Exec "daemon" "/usr/local/bin/hp-asmcli-temp.sh"
Exec "daemon" "/usr/local/bin/hp-asmcli-fans.sh"
Exec "daemon" "/usr/local/bin/hp-asmcli-pwr.sh"
Exec "daemon" "/usr/local/bin/zfs.sh"
</Plugin>
###Results :
hp-asmcli-temp.sh hp-asmcli-pwr.sh hp-asmcli-fan.sh
zfs.sh scrubing zfs.sh scrub remaining zfs.sh free space
It contains my json to import the graf query. You will need rename the hostname to your own.
Scrubing doesn't differentiate between a scrub and a resilver. It will report speed and remaining time. Might flag something else in the future. Also, ZFS ZED could be a nice source of info to monitor.