[CentOS] DL380g8 - smart array B320i - CentOS 6.4

[m.roth at 5-cent.us]

Rainer Duffner wrote:
>
> Am 01.07.2013 um 20:39 schrieb John R Pierce <pierce at hogranch.com>:
>
>> On 7/1/2013 11:30 AM, Nathan Duehr wrote:
>>> The significant problem we ran into was someone at an upstream vendor
>>> orders HP stuff via individual part numbers in a specific configuration
>>> for us, so we get a server, some disks, whatever... and assemble them
>>> on-site.  They didn't know (bad vendor, no donut) about the change or
>>> spaced it... and didn't send licenses... so you're sitting there with
>>> disks in a new server, all ready to load the OS as usual... and the OS
>>> can't find any disks.
>>
>> that sounds like a VAR problem.  if I'm buying from a VAR, I expect the
>> system to arrive as ordered and configured.
>>
>> As we buy direct from HP (big corp), I *ALWAYS* go through the entire
>> 'quickspec' page on any HP gear, carefully studying the options and
>> SKU's, any such licenses should be clear there.    For example, I
>> *always* get the full ILO license.
>
> Somebody correct me, but the B320 controller only comes in the "e"-type
> models of DL3x0 servers, right?
> We only order the "p" models and we generally don't need to enter licenses
> to access hard-drives, unless we want to create a RAID6…

And we like Dell, a lot. When we get a server with a PERC 6xx or 7xx, it
*works*, we don't need special licenses or whatever to activate it.
<snip>
> I'm not 100% sure, but I believe, with SuperMicro Servers, I'd  have to
> have a much larger (and better organized) inventory of spares that might
> only fit into specific age-group of servers…

Um, yes. We've got a ton of servers from Penguin, and in the last year,
we've decided we don't want any more, and will try to convince our folks
to not get them. They're *all* Supermicro, and over the years we've had a)
servers with hot-swap bays where one or more of the four *does* *not*
*work* (or the ones I've had to use a pencil to push wires that were
hanging down into the bay, preventing the drive from seating, which *is*
Penguin's fault); b) of the new 48 core and 64 core systems, we've sent a
significant percentage back for repair, mostly new m/bs. I don't trust
Supermicro's alleged "quality control", when we're buying high-end servers
and we have this problem.

          mark "wouldn't buy a SM m/b at home, either"