[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [cobalt-users] What cpu is in a cobalt raq4 - screaming lawsuits



On Sun, 3 Dec 2000, Carrie Bartkowiak wrote:

> Egads, Batman.
> Okay. Here's my reports, as following the various commands that have come
> through the list:
> 
> [root admin]# cat /proc/net/dev
> Inter-|   Receive                                                |  Transmit
>  face |bytes    packets errs drop fifo frame compressed multicast|bytes
> packe
> ts errs drop fifo colls carrier compressed
>     lo: 2491624   25626    0    0    0     0          0         0  2491624
> 256
> 26    0    0    0     0       0          0
>   eth0:131192296  866822    0    0    0     0          0         0 102737685
> 15
> 9123    0    0    0     0       0          0
> 
> >From this it would seem that I only have one nic.

Only one NIC is *configured and running*.

[snip]

> When typing in the 'dmesg' command in telnet, this is what I get:
> [root admin]# dmesg
> Linux version 2.2.14C11 (root@xxxxxxxxxxxxxxxxxx) (gcc version 2.95.2
> 19991024 (
> release)) #2 Wed Jun 28 00:55:51 PDT 2000
> Ignoring bogus EBDA pointer 3DDB000
> Detected 298806483 Hz processor.
> Calibrating delay loop... 596.38 BogoMIPS
> Memory: 127980k/131072k available (1076k kernel code, 416k reserved, 1540k
> data,
>  60k init)
> 
> <snipped>
> 
> Further down I only have etho0 (zero), no etho1 listed. As you can see from
> the screenshots, there's something weird going on in there too, got a bunch
> of lockups at the end.
> http://www.allaboutchoice.com/list/telnet1.jpg
> http://www.allaboutchoice.com/list/telnet2.jpg
> http://www.allaboutchoice.com/list/telnet3.jpg
> http://www.allaboutchoice.com/list/telnet3.jpg
> 
> I'm supposed to have a 15.3GB hdd, it's reporting as a 14.2BG; but that
> could be a matter of how the reporting mechanism gauges the size I suppose.

It depends how you define 1Gb. Some people (disk vendors) like to define
it as 1,000,000,000 bytes. If you check the size of the disk you'll see
that it is 14652*1024*1024 bytes which equals 15,363,735,552 bytes which
they of course round to 15.3Gb. I'm surprised they don't round it to
15.4Gb. I had a similar issue with IBM a couple of years ago. They ended
up giving me an 6 extra drives because they hadn't defined 1Gb as
1,000,000,000 bytes in their literature anywere.

> Ahh sh*t.  I could swear that my Phasi told me I had a 450. I just went to
> check it though, and it's telling me I have a 300:

With AMD, the jumper settings are important. It may well be an AMD 450CPU
which has been put in a board with the wrong jumper settings, clocking it 
at 300MHz. Doesn't help though, I know.

> CPU: 1 AuthenticAMD AMD-K6(tm) 3D processor 298 MHz Processor(s)
> Net Card info: 1 ethernet interface(s) eth0: Intel EtherExpress Pro 10/100
> at 0x1000, 00:10:E0:01:D2:8B, IRQ 11.
> IDE device(s) info: 1 IDE Device(s) hda: ST315323A, 14652MB w/512kB Cache,
> CHS=29770/16/63, (U)DMA
> PCI info: PCI bus devices: Host bridge: Acer Labs M1541 Aladdin V (rev 4).
> PCI bridge: Acer Labs M5243 AGP (rev 4). Non-VGA device: Acer Labs M7101 PMU
> (rev 0). ISA bridge: Acer Labs M1533 Aladdin IV (rev 195). IDE interface:
> Acer Labs M5229 TXpro (rev 193). Ethernet controller: Intel 82559ER (rev 9).
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Looks like your second ethernet card here.

> Maybe it's when hubby and I were shopping for servers, we were talking about
> so many different processors, 450 stuck in my head. Anyhow, I can make my
> Phasi report available if anyone would like to see it - although you have my
> word I just copied and pasted that stuff right out of it.
> 
> To CR's credit, on their offering they don't say specifically what processor
> speed you'll get, just an AMD x86. Here's the offer for a RaQ4i
> (non-enterprise, which is what I have):
>  RaQ 4i Platform
>  AMD X86
>  128MB DRAM
>  15.3GB Hard Disk
>  Includes ASP!
>  Includes PHP 4!
>  Includes MySQL!
>  Supports SSL
>  66 GB of Traffic
>  $0/Setup
>  $199/Month
> 
> Granted I get 50GB of bandwidth, that's the only thing that's changed since
> I signed up. Also, I suggested that they needed to take the MySQL off of
> there the first week I had the server and ended up installing MySQL myself
> because it wasn't on the machine... but they haven't done that.
> 
> On the Cobalt data sheet for the RaQ4 it says this:
> Ordering information: Part Number R46 CE2 GIU
> RAQ 4i with 128MB, DRAM Single 15GB IDE disk, dual ethernet, SCSI, PCI slot
> Note that it doesn't say what processor.
> Further up, in Hardware, it says
> "450 MHz Intel compatible processor"
> *Not* that it has to be *genuine* Intel - just Intel compatible.  Although
> finding out there's a 300 where my 450 should be is still disappointing. It
> also says that there are to be dual 10/100 Base-T ethernet interfaces. Looks
> like I only have one.

I think the second one's just not configured.

> I know that I've seen other cobalt dedicated hosts put 'AMD x86' on their
> details page. So could this be something that Cobalt is offering now, but it
> wasn't in the plans when the data sheet was printed?
> 
> Is it possible that this is not CR's purposeful misleading of customers, but
> rather that they were shipped the wrong configuration from Cobalt itself?
> I've seen quite a few messages this past week of people that have had to RMA
> a lot of RaQs and are getting out of Cobalt all together because of it.
> 
> You can go to CR's site and look at the pictures of their July shipment and
> their October shipment, which was about ten times the size. I just can't
> believe that they would even attempt to take the time to take 450 processors
> out of all of those RaQs and slip in 300's instead. A mistake had to have
> been made somewhere. At this point we don't know; it could be a mistake that
> generated from Cobalt's end; shipping the wrong machines or even
> purposefully packaging a 300 where there was supposed to be a 450 and
> shipping it to CR. We don't know.

Could just be that the manufacturing line have put the CPUs into boards
with the wrong/old jumper configiration as I mentioned above.

> Maybe we should ask the Vice President of Cobalt Sales - he was there at
> CR's NOC (see pictures on the site).
> 
> I've also got a bunch of emails with trace routes from when things got slow
> or there was no connection. (I always include a tracert if there is a
> problem so that Sam can track it down.)
> 
> Sigh.
> It is looking like, even if CR is found to be innocent of what everyone's
> claiming, we'll all have to find a new NOC anyway because of this lawsuit
> and everything it's going to cost them.

You Americans do love your lawsuits.

> Double sigh.
> Carrie Bartkowiak

regards,
Colin.