Start a Conversation

Unsolved

This post is more than 5 years old

26997

July 30th, 2014 12:00

BLADE CHASSIS1 PowerEdge M1000 e

​This M1000e has 8 ESXi 5.1 hosts and directly connected to the Dell Compellent SAN. ​

​We need to change the mid-plane of the M1000 e.​

​We will shut down all the 8 Blades. Do I need to do anything to disconnect the M1000 e from the Compellent​

​How do I go by shutting down the M1000e for repair.​

​Thanks​

Moderator

 • 

8.5K Posts

July 30th, 2014 14:00

Sweetmail1982, 

With you bringing the Blades offline first you are following the procedure for disconnecting from the Compellent SAN. Then once the Blades are off then you can simply hold down the power button on the M1000e control panel to power down the enclosure. Once you are done with the repair then use that to power on the M1000e and then after that then bring up the Blades. The Compellent SAN will be continuously searching, so once the Blades are online it should connect on its own.

Let me know if this helps to answer your questions.

July 30th, 2014 15:00

Thank you. 
 

What if we need to unplug the ISCS cable off the Compellent SAN from the chassis. Would it upset the compellent SAN?

Moderator

 • 

8.5K Posts

July 31st, 2014 06:00

The only issue you may run into is if the cable is reinstalled to the exact same ports. So long as you mark your ports and restore it there you should be fine. Let me know how it goes.

Also, why are you needing to replace the midplane on the enclosure, just curious?

July 31st, 2014 06:00

Apparently after all the money we have spend on setting up an enterprise virtual environment, we have a single point failure. We only have the Dell Blade Chassis.

A month or so into the new equipment, we ran into networking issue. All of our hosts complaint about loosing some networking capability. Dell told us to re-sit one of the network card of the chassis. We did and everything was fine. Then it happens a few more times and everything was fine after the re-sitting of the I/O Module A2. 

When the network card module went crazy, the whole VM environment will freeze. As soon as the card get resetted, Everything starts working like nothing has happen. 

The last time when we experienced the problem, we request a new card assuming the original card got a problem.

When the new card came in, we put it in, all the VMs lost network. As soon as we pull the new network out of the Chassis, everything works again. Of course, I lost a nic on each of the hosts.

Dell then send another brand new network card for replacement, no luck. it did not work.

Then they guess that it is the mid-plane. Now we have to shut down 90% of the VMs for hours. It is not good. 

I don't understand whey Dell doesn't give us an loaner and work with us to as least keep some of mission critical machines up. Like public safety servers.

This is the story and again, they don't have any log or anything to really tell them what is wrong.

Anything suggestion or advise is most welcome.

Thanks

Moderator

 • 

8.5K Posts

July 31st, 2014 07:00

Sweetmail1982,

What is the network module we are referring to?  Also, if you access the CMC, is there anything showing in its log?

July 31st, 2014 08:00

The log only say IO module A2 is removed and IO module A2 is online everytime we put the card in and out... the name is MXL 10/40GbE. Can you see the image. Want to show you the screen shot of my chassis and output of the hardware log.[View:~/cfs-file.ashx/__key/communityserver-discussions-components-files/956/chassis.docx:550:0][View:~/cfs-file.ashx/__key/communityserver-discussions-components-files/956/A2.pdf:550:0]

Moderator

 • 

8.5K Posts

July 31st, 2014 09:00

Before replacing the midplane. What is the firmware revision on the MXL 10/40, is it on or before 8.3.16.1?

July 31st, 2014 09:00

how do you find out? Sorry . don't know how

Moderator

 • 

8.5K Posts

July 31st, 2014 10:00

You can Putty  into the IP address on the MXL switch through the network. Then Show Version on the MXL command line. 

July 31st, 2014 11:00

Sorry

I was not available earlier. Checking it now

Moderator

 • 

8.5K Posts

July 31st, 2014 11:00

Were you able to locate the version?

July 31st, 2014 12:00

The broken network switch is laying on my desk. I assume the other one is identical. Here is the print out of "show version". Thanks

[View:~/cfs-file.ashx/__key/communityserver-discussions-components-files/956/switcha1.docx:550:0]

Moderator

 • 

8.5K Posts

July 31st, 2014 12:00

Thanks Sweet,

Could you do me a favor, prior to swapping the midplane. Would you follow this link, login, and find the firmware update for the MXL, I believe it is at 9.4 or 9.5 currently. - https://www.force10networks.com/csportal20/Main/SupportMain.aspx
I
t sounds like this might be what is causing this. If so then verify by testing the original switch as well. If you don't have access to the page there is a Access Request link on the left. 

Let me know what you see.

July 31st, 2014 12:00

ok, thank you,

Are you saying that, may the MXL that is causing all this problem may just be ir has different firmware then the one that is left in the Chassis. 

If I would have put the original MXL back in and check the firmware level, both firmware level should be different and I need to make them the same?

If that is the case, the midplane maybe good and no need to be replaced?

Moderator

 • 

8.5K Posts

July 31st, 2014 12:00

I am not sure where the other module is at in revision, but if it is 9.2 or older then it also is behind as well. I wanted to get them current to verify that is the case, BEFORE we blame the midplane and causing a long downtime. Just want to cover the bases.

No Events found!

Top