RAID monitoring tools

[Mike Holloway]

Greg, you might find this thread useful:

http://lists.freebsd.org/pipermail/freebsd-proliant/2006-February/ 
000058.html

It contains a perl script to run hpasmcli and exit if it doesn't  
return in a reasonable amount of time.  John also mentioned disabling  
ASR on the server in the post below, which seemed to fix my reboot  
problem.  Unfortunately, even still, the hpasmcli utility on FreeBSD  
is unstable for me in it's current form.  You may have different  
results.

http://lists.freebsd.org/pipermail/freebsd-proliant/2006-February/ 
000059.html


-mike



On Oct 31, 2006, at 8:17 AM, Greg Eden wrote:

>
> Hi there
>
> Thanks for this very useful post to the proliant list. i'm just in  
> the process of getting to grips with SNMP and nagios with my  
> proliant hardware running freebsd - this will help no end.
>
> another task i'm just starting on is using output from SHOW TEMP,  
> SHOW FANS from the hpasmd/hpasmcli tools to graph hardware info  
> with nrg. if i get anywhere with it i'll post it to the list.
>
> cheers
>
> greg.
>
>
>
>> Whoops, meant to copy the list...
>>
>>
>> Appreciate the pointer to camcontrol, I previously had just been
>> using swatch to watch syslog and send messages to nagios via nsca.
>> My problem was that I never knew the initial state of the disks until
>> an event happened in syslog.
>>
>>
>> For reference, here's what I saw from camcontrol during my tests
>> (FreeBSD 6.0 rel):
>>
>> During normal operation of the raid:
>> # camcontrol inquiry da0 -D
>> pass0: <COMPAQ RAID 1  VOLUME OK> Fixed Direct Access SCSI-0 device
>>
>> After removing one of the raid member disks:
>> # camcontrol inquiry da0 -D
>> pass0: <COMPAQ RAID 1  VOLUME inte> Fixed Direct Access SCSI-0 device
>>
>> After re-inserting the raid member disk:
>> # camcontrol inquiry da0 -D
>> pass0: <COMPAQ RAID 1  VOLUME reco> Fixed Direct Access SCSI-0 device
>>
>> And about 45 minutes later:
>> # camcontrol inquiry da0 -D
>> pass0: <COMPAQ RAID 1  VOLUME OK> Fixed Direct Access SCSI-0 device
>>
>>
>>
>> And here's the configuration I use for swatch to feed nsca in  
>> realtime:
>>
>> watchfor   /ciss0.*removed/
>>          exec "/usr/local/bin/nsca_report 2 \"Disk Array\" Hot-plug
>> drive removed"
>>
>> watchfor   /ciss0.*failure/
>>          exec "/usr/local/bin/nsca_report 2 \"Disk Array\" Physical
>> drive failure"
>>
>> watchfor   /ciss0.*inserted/
>>          exec "/usr/local/bin/nsca_report 1 \"Disk Array\" Hot-plug
>> drive inserted"
>>
>> watchfor   /ciss0.*recovery->recovering/
>>          exec "/usr/local/bin/nsca_report 1 \"Disk Array\" Drive is
>> rebuilding..."
>>
>> watchfor   /ciss0.*recovering->OK/
>>          exec "/usr/local/bin/nsca_report 0 \"Disk Array\" Drive has
>> successfully rebuilt."
>>
>>
>>
>> For completeness, here's the nsca_report script that I use to send
>> the alarms to nagios, substitute your own thishost and -H:
>>
>> #!/bin/bash
>>
>> outcode=$1
>> thisservice=$2
>>
>> thishost=`echo $HOSTNAME | sed -e "s/\./ /g" | cut -f 1 -d ' '`
>>
>> shift
>> shift
>>
>> echo -e "${thishost}\t${thisservice}\t${outcode}\t$*\n" | /usr/local/
>> bin/send_nsca -H www -c /usr/local/etc/send_nsca.cfg 2>&1 >> /dev/ 
>> null
>>
>>
>>
>> -mike