check_openmanage and net-snmp v3
Verhaeghe, Koen
Koen.Verhaeghe at meucci-solutions.com
Fri Feb 12 10:28:26 CET 2010
Hi Trond,
Thanks for your reply and the new script.
These are the results:
With windows SNMP (v2) it works:
./check_openmanagetest -H xx.xx.xx.xx -C xxxxx -p multiline -t 120 -o 3
-b ctrl_fw=all/ctrl_driver=all/ctrl_stdr=all
OK - System: 'PowerEdge 2950', SN: '7Z3DR3J', hardware working fine, 1
logical drives, 2 physical drives
----- BIOS='2.2.6 02/05/2008', DRAC5='1.33', BMC='2.05'
----- Ctrl 0 [PERC 5/i Integrated]: Fw='5.2.1-0067', Dr='1.21.00.32',
Storport: '5.2.3790.3959'
----- Encl 0:0:0 [Backplane]: Fw='1.05'
----- OpenManage Server Administrator (OMSA) version:
'6.2.0'|'fan_1_system_board_fan_1_rpm'=5175RPM;0;0
'fan_2_system_board_fan_2_rpm'=5250RPM;0;0
'fan_3_system_board_fan_3_rpm'=5250RPM;0;0
'fan_4_system_board_fan_4_rpm'=5175RPM;0;0
'pwr_mon_0_ps_1_current_1'=0.4A;0;0
'pwr_mon_1_ps_2_current_2'=0.4A;0;0
'pwr_mon_2_system_board_system_level'=184W;916;964
'temp_0_system_board_ambient'=16C;42;47
With net-snmp v3 (version 5.4.2.1) on the same server, diabling the
windows snmp, I get:
./check_openmanagetest -H xx.xx.xx.xx -P 3 --authprotocol md5 -U xxxxxx
--authpassword xxxxxxx --privpassword xxxxxx --privprotocol des -p
multiline -t 120 -o 3 -b ctrl_fw=all/ctrl_driver=all/ctrl_stdr=all
SNMP ERROR [processors]: Received genError(5) error-status at
error-index 3.
This normally indicates a too low version of OMSA, but I am using 6.2.0.
Any ideas?
Regards,
Koen
-----Original Message-----
From: Trond Hasle Amundsen [mailto:t.h.amundsen at usit.uio.no]
Sent: 2010-02-11 16:46
To: Verhaeghe, Koen
Cc: nagios-users at lists.sourceforge.net
Subject: Re: [Nagios-users] check_openmanage and net-snmp v3
"Verhaeghe, Koen" <Koen.Verhaeghe at meucci-solutions.com> writes:
> Hi All,
>
> does anyone have an explanation for this:
> when using check_openmanage with snmp v3, the script exits because
> some OIDs do not exist for a type of server.
> (e.g. '1.3.6.1.4.1.674.10893.1.20.130.4.1.9' =>
'arrayDiskEnclosureID'
> for PowerEdge 860).
>
> output:
> ./check_openmanage -H xx.xx.xx.xx -P 3 --authprotocol md5 -U xxxx
> --authpassword xxxxx --privpassword xxxxx --privprotocol des -p
> multiline -t 120 -o 3 -b ctrl_fw=all/ctrl_driver=all/ctrl_stdr=all
> SNMP ERROR [storage / pdisk]: The requested entries are empty or do
> not exist.
>
> When enabling the windows snmp service again and disabling the
> net-snmp v3, I get the correct output:
>
> ./check_openmanage -H xx.xx.xx.xx -P 2 -C xxxxxx -p multiline -t
> 120 -o 3 -b ctrl_fw=all/ctrl_driver=all/ctrl_stdr=all
> Physical Disk 0:1 [Ata WDC WD800JD-75MSA3, 0GB] on ctrl 0 needs
> attention: Failed
> Logical drive 0 'Windows Disk 0' [RAID-1, 73.57 GB] on ctrl 0 needs
> attention: Degraded|'fan_1_bmc_cpu#fan'=3225RPM;0;0
> 'fan_2_bmc_dimm_fan'=3150RPM;0;0
> 'temp_0_bmc_planar'=31C;48;53
>
> tested with:
> OMSA version: 5.1 and 6.2
> Net-snmp (x86) versions 5.4.2.1 and 5.5 NET::SNMP 6.0.0 on the nagios
> server
>
> Any ideas?
> I've tried commenting out the OIDs that do not exist (and all related
> script steps) but then the output gives 'OK', but I know there is a
> degraded disk...
>
> ./check_openmanage -H xx.xx.xx.xx -P 3 --authprotocol md5 -U xxxx
> --authpassword xxxxx --privpassword xxxxx --privprotocol des -p
> multiline -t 120 -o 3 -b ctrl_fw=all/ctrl_driver=all/ctrl_stdr=all
> OK - System: 'PowerEdge 860', SN: 'J478F3J', hardware working fine, 1
> logical drives, 2 physical drives
> ----- BIOS='A05 10/04/2007', DRAC4='1.60', BMC='1.75'
> ----- Ctrl 0 [SAS 5/iR Adapter]: Fw='00.10.51.00.06.12.05.00',
> Dr='1.21.08.00'
> ----- OpenManage Server Administrator (OMSA) version:
> '5.1.0'|'temp_0_bmc_planar'=30C;48;53
>
> On other types of servers I get a similar error for [cooling] (e.g on
> a
> 2950)
Hi Koen,
I'm the author of that plugin. To be honest, I've never actually tested
the SNMPv3 stuff. I just pass the options to Net::SNMP and let it handle
it, and hope that it works. You are the first to report SNMPv3 troubles,
and I assume that the SNMPv3 users are a minority.
I'm always interested in fixing bugs, but I'm unable to reproduce this
problem. I see that you're checking a Windows box. I have none of those
to play with, but I have set up SNMPv3 on a RHEL5 box. Checking the
RHEL5 host via SNMPv3 works just fine:
$ ./check_openmanage -H myhost -P 3 --authprotocol md5 -U xxxx \
--authpassword xxxx --privpassword xxxx --privprotocol des
Controller 0 [SAS 6/iR Integrated]: Driver '3.04.07rh' is out of date
Windows + OMSA + SNMP has had some problems in the past, but at least
for SNMPv2c and SNMPv1 these issues should be resolved with OMSA 5.5.0.1
and later versions. It seems there are still issues with SNMPv3.
In the past, there have been problems with SNMP and using the Net::SNMP
function get_entries() vs. get_table(). The former is preferred because
it is faster, since we're not interested in all the OIDs. This is
especially true for servers with many physical disks.
I have created a test version that fetches the cooling OIDs with
get_table() instead of get_entries() if SNMPv3 is used. This version is
available here:
http://folk.uio.no/trondham/tmp/check_openmanage-snmpv3test
Can you try this version on the servers where checking the cooling
devices fail?
(It's a bit more complicated for physical drives).
PS. Please upgrade to OMSA version 5.5.0.1 or later. Previous versions
are known to perform badly with SNMP on Windows.
Cheers,
--
Trond H. Amundsen <t.h.amundsen at usit.uio.no> Center for Information
Technology Services, University of Oslo
------------------------------------------------------------------------------
SOLARIS 10 is the OS for Data Centers - provides features such as DTrace,
Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW
http://p.sf.net/sfu/solaris-dev2dev
_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue.
::: Messages without supporting info will risk being sent to /dev/null
More information about the Users
mailing list