On 11/09/2017 09:08 AM, Lamar Owen wrote:
> On 11/08/2017 06:25 PM, ToddAndMargo wrote:
>> On 11/07/2017 06:22 PM, Nico Kadel-Garcia wrote:
>>> On Tue, Nov 7, 2017 at 8:43 PM, ToddAndMargo <[log in to unmask]>
>>> wrote:
>>
>>>> And you know the Cxxx series of chipsets have been around
>>>> for a while now. Just not long enough to be out of
>>>> production at which point it will appear on Red Hat
>>>> compatibility list.
>>>
>>> Nonsense. Our friends over at Red Hat are continuously supporting
>>> leading edge *server* hardware.
>>
>> Niko! The C236 chipset *IS* a server grade chipset!
>> And it has been around for a long time. No doubt Red Hat
>> will eventually support it in about five years, which is
>> typical of them and useless to me.
>>
>
> According to SuperMicro's support matrix for their C236 motherboards at
> http://www.supermicro.com/support/resources/OS/C236.cfm most of their
> C236 motherboards are supported, with some caveats, by RHEL7 (and thus
> CentOS or SL). So the C236 chipset itself is supported by EL7, and has
> been for a while.
>
> But it just so happens that the particular C236-chipset board you used,
> the X11SAE-M, is apparently supported by SuperMicro for EL 7.1, but not
> later. Now, others with a C236, such as the X11SAT and X11SAT-F, list
> support for 7.2 and 7.3 (the chart hasn't been updated for 7.4). So you
> bought a motherboard that, for EL7.4, is not supported with that OS by
> SUPERMICRO. Not Red Hat, but SuperMicro, the manufacturer. To be
> blunt, but not intended to be harsh, if you want to use EL, do you
> homework and buy a supported board. (Google-fu: search for the string
> 'C236 supermicro support matrix' and this chart is third on the list,
> front page).
>
> I personally have not had that much problem using newer software in
> CentOS 7, thanks to the Software Collections concept and repositories. A
> few things are older, but the latest Chrome works fine, the latest ESR
> Firefox and Thunderbird both work fine, etc. The kernel continues to get
> new hardware drivers backported to it, and it has been pretty stable for
> me (I'm having an issue at the moment with the nVidia binary drivers
> packaged by ELrepo, but that's not a Red Hat thing).
>
> SuperMicro typically has great support for all things Red Hat, but this
> is one of those cases where it just isn't supported; if they could
> support it they would (they being SuperMicro, not Red Hat, since
> SuperMicro has the resources to do that support). Safer bet is buying a
> Dell, HP, or other server that has as one of its options Red Hat
> Enterprise Linux; Dell especially is very good about this.
Hi Lamar,
I find the hardware compatibility charts to be so wildly out
of date that they are next to useless. And the problem
occurred across all versions of 7.x that I tested.
I have seen those charts used to "Oops, not on our chart.
Your problem, not ours". Supermicro's memory charts
terrible for this. You can't find any of the memory they
certify as it is all out of production. "Memory
problems? Not on our chart!"
Also, Red Hat made no complaint about this not being on their
chart, so they expected it to work. Red Hat L-O-V-E-S
to "WON'T FIX" bug reports. That would have been an easy
one for them to worm out of. Instead they pulled the
"we don't have the hardware" when it could have easily been made
available tot them from Supermicro. They still wormed
out of it, they just had to be a bit more creative.
I also asked Supermicro to remove the X11SAE-M from their Red
Hat table. I don't know if they did or not.
If you go and read the bug report I made with Red Hat, you will
find that the X11SAE-M worked perfectly when run from
the EXACT same Live USB that I installed it with. This is
a timing issue. The Live USB runs slower than native
hard drive.
I though I done complete do diligence. It worked on the Live USB
PERFECTLY. It only cost me a bit over $1000 in free consulting
to figure it out. I had to completely rip SL 7.x out and replace
everything with Fedora on a running production server FREE OF
CHARGE. Yup, it did occasionally boot on 7.x. About one out
of three times (it got worse). If SL did manage to boot,
it worked perfectly. Yes, I am still pissed.
RHEL and Clones are great for set and forget (appliance)
applications (if they work initially), but not where on
going innovation is occurring.
And I firmly believe that there is a difference between
assuring stability and procrastination.
-T
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Computers are like air conditioners.
They malfunction when you open windows
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
