1 |
On 10/31/06, Duncan <1i5t5.duncan@×××.net> wrote: |
2 |
> |
3 |
> "Vesna Petrovic" <vesna.petrovic@×××××.com> posted |
4 |
> 60bedadc0610311046q23d77c5fu671c7330a2f09a14@××××××××××.com, excerpted |
5 |
> below, on Tue, 31 Oct 2006 13:46:29 -0500: |
6 |
> |
7 |
> > The system has 2 AMD Opeteron Processors 252, 5 disks - 1IDE Maxtor |
8 |
> > 6B200R0 and 4 SCSI Maxtor 6L300S0, and probably irrelevant ATAPI 48X |
9 |
> DVD-ROM |
10 |
> > DVD-R CD-R/RW drive, Ethernet controller: Broadcom Corporation NetXtreme |
11 |
> > BCM5703X Gigabit Ethernet (rev 02), RAID bus controller: Silicon Image, |
12 |
> Inc. |
13 |
> > SiI 3114 [SATALink/SATARaid] Serial ATA Controller (rev 02), FireWire |
14 |
> (IEEE |
15 |
> > 1394): Texas Instruments TSB43AB22/A IEEE-1394a-2000 Controller |
16 |
> (PHY/Link), |
17 |
> > VGA compatible controller: ATI Technologies Inc RV350 AP [Radeon 9600] |
18 |
> > 05:00.1, Display controller: ATI Technologies Inc RV350 AP [Radeon 9600] |
19 |
> > (Secondary). |
20 |
> > Kernel version is 2.6.17 built using gentoo-sources. |
21 |
> |
22 |
> I have but a few possibilities to suggest, but I can say I've been using a |
23 |
> similar system since late 2003, upgrading it over time, and know it works. |
24 |
> FWIW, Tyan s2885 mobo here, dual Opteron 242s, four-way SATA RAID on SiI |
25 |
> 3114 SATA controller, 8 gig memory, Firewire, Radeon 9200 (so the ATI r200 |
26 |
> series not the r300 series you have, mine is AGP), a DVDRW, CDRW, and |
27 |
> older |
28 |
> PATA hard drive on PATA, etc. |
29 |
> |
30 |
> Possibilities: |
31 |
> |
32 |
> I had bum memory for awhile. Twas a mess, until I got a BIOS upgrade that |
33 |
> allowed me to declock it a notch (to pc3000 level from the pc3200 it was |
34 |
> rated). A memory upgrade cured that. |
35 |
> |
36 |
> The usual possible hardware suspects including the power supply train |
37 |
> (from the one in the computer itself, to what's coming in on the line, to |
38 |
> the UPS if any), and possibly overclocked CPUs. With all those hard |
39 |
> drives, it's possible you are simply underpowered. I think I'm running a |
40 |
> 550 watt rated Vantec that's rated to 650-ish spike. |
41 |
> |
42 |
> At one point there was an issue with the firewire driver and x86_64 SMP. |
43 |
> Since I'm not using Firewire for anything here, IIRC I disabled it in BIOS |
44 |
> as well as the kernel driver for it. I've not followed up but I'd /guess/ |
45 |
> the problem is fixed by now. Still, it's worth trying that, and disabling |
46 |
> any other stuff (like USB possibly in a server environment, which it seems |
47 |
> you are in) you don't actually use. |
48 |
> |
49 |
> Those are all shots in the dark, but this is mainly to say that what I |
50 |
> have here seems to work very well, at least since I fixed the memory |
51 |
> issue, which was simply generic memory not worth its rating and an |
52 |
> inability at the time to declock it in BIOS. |
53 |
> |
54 |
> -- |
55 |
> Duncan - List replies preferred. No HTML msgs. |
56 |
> "Every nonfree program has a lord, a master -- |
57 |
> and if you use the program, he is your master." Richard Stallman |
58 |
> |
59 |
> -- |
60 |
> gentoo-amd64@g.o mailing list |
61 |
> |
62 |
> |
63 |
Bob, Duncan, thank you. |
64 |
You were right, the problem does seem to be memory. I swapped the memory |
65 |
modules, and the system was not freezing any more. I started getting |
66 |
'Segmentation fault' errors during compile, so I limited the memory clock in |
67 |
BIOS @ 166MHZ. Now everything works fine. |
68 |
How can I find out what is the best value to set the memory clock limit to? |
69 |
I tried 200MHz, and the segmentation faults still occurred. When I set 183 |
70 |
MHz, BIOS reported that the memory clock is 'unknown'. So I set it to |
71 |
166MHz, which was the next value in the list. Is this a good method for |
72 |
determining the best value? |
73 |
|
74 |
Vesna |