Cisco B200 M3 Blade server discovery failure handling

Customer blade Server Upgrade, purchased two Cisco B200 M3, after arrival we looked at the version of B200 M3, and then the UCS manager upgrade, but after the upgrade found that the two blades in the discovery process, only to go to 7% failed! As shown in the following:

650) this.width=650; "src=" Http://s1.51cto.com/wyfs02/M02/79/32/wKioL1aLbh3yCJ_tAAVW-PfVfyE410.jpg "title=" Zfedu01.jpg "alt=" Wkiol1albh3ycj_taavw-pfvfye410.jpg "/>

Also, view the alarm information and Cisco documentation as follows:

650) this.width=650; "src=" http://s4.51cto.com/wyfs02/M01/79/32/wKioL1aLbquwcyXkAAHSB9DvMCg874.jpg "style=" float: none; "title=" zfedu03.jpg "alt=" Wkiol1albquwcyxkaahsb9dvmcg874.jpg "/>

650) this.width=650; "src=" http://s5.51cto.com/wyfs02/M02/79/34/wKiom1aLboeSPVymAAG7vcN3438327.jpg "style=" float: none; "title=" zfedu02.jpg "alt=" Wkiom1alboespvymaag7vcn3438327.jpg "/>

The document you are looking at does not get useful information. Where exactly is the problem? For further verification, we have done the following:

1. Remove the blade server and install only one CPU and one memory. In this state the server can discovery through.

2, install two CPUs and two memory, at this time the phenomenon is still, still can not pass.

3, will CPU1 and CPU2 swap, CPU2 is separate orders, CPU1 is with the blade come over, only install CPU2 and a memory, discovery process can pass.

4, the original one B200 M3 memory all removed and the existing memory swap, installation of two cpu,discovery process or can not pass.

5, only in the CPU2 slot to install a CPU and a memory, still not (later learned that in the case of only one CPU, can only be inserted in CPU1 slot 1)

6, replace the blade slot, also not ...

7, Recover, also not ...

..................

Tried all sorts of still not, the preliminary judgment problem may be out of the motherboard, so with this question to Cisco opened a case, then Cisco TAC engineers collected a log, according to the log to determine the problem may also be on the motherboard. Finally helped us to do an RMA, re-send two pieces of motherboard come over!

Look forward to always beautiful, but that is only looking forward to!

To the two of the motherboard to the RMA arrived, we put the CPU, memory and other components plugged in, insert the knife box, the hell of things happen again, the discovery process is still walking to 7% will not move! Both of them!

The feeling is not common sense, but the problem still appears!

We will appear the problem reported to the big boss, Big Boss also deeply puzzled, impossible to send is bad ah! So the New Year's day after the first boss to visit the scene.

Big Boss through a check, found that the two blades of the boardcontroller of the firmware not show version display, that is not see what version, is not because of this caused the problem? The UCS Manager is then logged in via SSH, and the following actions are performed:

c6248a-a# Scope Chassis 2
C6248a-a/chassis # scope Server 6
C6248a-a/chassis/server # Scope Boardcontroller
C6248a-a/chassis/server/boardcontroller # Activate firmware 14.0
Warning:when committed this command would reset the end-point
C6248a-a/chassis/server/boardcontroller # Commit-buffer

When the blade is found again, it passed the 7%, until the discovery of success! At this point, the RMA came over the two blades of the problem solved!

Big Boss, wow Quack ... (*^*)

After two blades are normal, are we thinking that the other two are also for this reason? So we re-toss, the original blade server inserted into the knife box again for inspection. However the matter is willing to violate, finally had to send out the faulty blade!

**********************************************************************************************

Summarize:

1, the fault treatment must be careful, can not let go of every link.

2, do not use inertial thinking to treat every fault, the pig is hit the dead tree, but the rabbit is not necessarily hit the dead tree. My colleague and I made this mistake on the blade of the RMA.

3, to deal with the failure of the matter to the smooth! A little bit, step by step!

This article is from the "Xunil" blog, make sure to keep this source http://136464.blog.51cto.com/126464/1731742

Cisco B200 M3 Blade server discovery failure handling