point of failure on c3000

[jkimes]

I am trying to find out if anyone has experienced a failure on a HP c3000, was told that it was the back plane only to find out it was the OA. As i get ready to add more balde serbers, I would like to know if there are any other known points of failure in the c3000.

[The Brit]

I assume you are refering to "Single points of Failure", in which case there can be many, depending on what you put in there for IC Modules, Blades, Mezz cards and HBA's.

Remember, the c3000 is designed for economy, not for redundancy. For example, all of your blade LOM NIC's are hard mapped to Interconnect Bay 1. If the module in Bay 1 fails....

(you can get around this by teaming with a NIC port on a MEZZ card, but this requires you to study the Port Mappings and configure accordingly.)

[jkimes]

Spell checker not withstanding, thank you The Brit for your for your quick response. Just trying to prove to management that the OA failure was "out of the ordinary" and not a common occurence.

[gregersenj]

OA failure is rare. But it will not take down running servers.

The backplane or more correct the midplane, has no active components.
There's a few components on the midplane, that hold the chassis s/n and model no and pdu type. This is for house keeping information only.
The only problem on the midplane, is damage to connectors, due to human error.
This will caurse a bay to become unuseable, and you will have to shut down all servers to replace the midplane.

But agree, you do have a single point of failure on inteconnect bays 1 and 2.

I have seen once, where the communication between the OA and the VC module in Interconnectbay 1 was lost, and we had to reseat the VC module. This did caurse a few minutes outage on the network.

BR
/jag

[gregersenj]

Interconnect bays 3 and 4, aren't redundant either.

/jag

[jkimes]

Thanks gregerseni! Good points, no pun intended!