/collectd_scripts

Scripts for collectd : HP Proliant G6, ping, ZFS

Primary LanguageShellGNU General Public License v3.0GPL-3.0

collectd_scripts

Scripts for collectd : HP Proliant G6, ping, ZFS

hp-asmcli-*

Uses hpasmcli to read : Fans, Power usage (power meter in PSU), Temperatures (22 readings)

hp-ssacli-array.sh

To monitor the array, but work in progress (stand-still).

ping.sh

You can ping a few hosts with this script. The build-in ping doesn't match the system pings.

zfs.sh

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 alt text hp-asmcli-pwr.sh alt text hp-asmcli-fan.sh alt text

ping.sh alt text

zfs.sh scrubing alt text zfs.sh scrub remaining alt text zfs.sh free space alt text

The folder grafana

It contains my json to import the graf query. You will need rename the hostname to your own.

Notes :

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.