Start a Conversation

This post is more than 5 years old

Solved!

Go to Solution

4569

February 22nd, 2017 08:00

Monitoring M1000e chassis. How? IPMI? SNMP?

Hello!

How do you people do hw-monitoring of the M100e chassis?

A blade can be checked via IPMI from the iDRAC. Or via the openmanage package, but i would prefer to be platform-agnostic here.

I cant see that any of the sensors from the drac-ipmi-system references any PSUs or similar.

The PSUs are in the chassis. So how are we supposed to monitor that? :)

I have looked and looked for any "enable IPMI" in the CMC of the chassis, but cant find it.

Any suggestions?

  wbr / Alex

1 Rookie

 • 

93 Posts

April 26th, 2017 02:00

I actually found a nice command to use in the CMC. Its quite easy to automatize this.

On your surveillanve-server, create a ssh key-pair.

Then upload the public key in the CMC like this

$ pkey=`cat .ssh/id_rsa.pub | tr -d '\n'`       # gives the variable 'pkey' the contents of the public key

$ ssh root@CHASSIS "racadm sshpkauth -i svcacct -k 1 -p 0xfff -t '$pkey'"    # will ask you for the CMC root pwd (calvin by default)

The thing is, i didnt see how to set a ssh key for the root account, so you set the key for an account called "service".

Then you should be able to login to the CMC as user service, like this

$ ssh service@CHASSIS    # make sure you use the ssh key now

And FINALLY, you can run the command "getactiveerrors"

$ getactiveerrors
There are no messages.

SO, this is easy to script, for instance to use with nagios/nrpe, like this (you can also specify in .ssh/config that you want to use the user 'service' when you connect to that host)

#!/usr/bin/env bash

chassis_name=$1

if [ -z "$chassis_name" ]
then
    echo "Usage: $0 "
    exit 2
fi

out=`ssh service@$chassis_name getactiveerrors | tr '\n' ',' | tr -s ' -'`
r=$?

if ((r))
then
    echo "Failed to ssh to $chassis_name."
    exit 2
fi

if [ "${out:0:22}" == "There are no messages." ]
then
    echo "Chassis $chassis_name all OK."
    exit 0
else
    echo $out
    exit 2
fi

1 Rookie

 • 

93 Posts

March 7th, 2017 08:00

Surely someone must be monitoring the M1000e hardware somehow? How?? :)

1 Rookie

 • 

53 Posts

March 7th, 2017 15:00

What are you using to monitor? If its nagios, you can made a script to parse output from dmsa.

 /opt/dell/srvadmin/bin/omreport storage controller controller=1 info=pdslotreport

1 Rookie

 • 

93 Posts

March 7th, 2017 23:00

What are you using to monitor? If its nagios, you can made a script to parse output from dmsa.

 /opt/dell/srvadmin/bin/omreport storage controller controller=1 info=pdslotreport

Perhaps you are thinking about an MD1000 disc-enclosure here?

I am talking about the M1000e blade server chassis.

1 Message

April 20th, 2017 14:00

I'd be interested in knowing too! I can't find any information on remote monitoring for M1000e blade chassis. There is Dell OpenManage Essentials, but that hasn't been updated since 2012 and I can't get it to detect the CMCs on our M1000e enclosures anyway.

1 Message

April 25th, 2017 22:00

Im trying to use PA Server Monitor but am unable to get it working... Im using IPMI (what they recommend to use) but i keep getting the following error: "Communication error while connected to Baseboard Management Controller (BMC)  (err -16)"

Have you had any luck? 

1 Rookie

 • 

93 Posts

April 26th, 2017 02:00

Well, it simply looks like IPMI is not available from the chassis, right? :)

No Events found!

Top