Interesting. Did you use 86+ or 86? 86+ is the most up-to-date. I'm surprised every server got that error, though I've never personally used memtest on a server box.
On Sat, May 8, 2010 at 9:18 AM, Frederic Breitwieser < [email protected]> wrote: > Cc2iscooL: > > >>Well, you do keep getting memory errors, I know you've said you've > >>switched out sticks, but have you actually tried any sort of diagnostic > program yet? > > Upon your suggestion, I can now say yes, I have. Using memtest86 off a > boot > floppy, I get a memory error right away, and it repeats for each test. I > swapped out the memory again with the old ones, and got a memory error > right > away, in the exact same spot. > > I powered down another server, yanked it's memory, shoved it in, got a > memory error again in the exact same spot. > > So, I went on a memory-swapping, memory testing bonanza on all 16 servers, > since the time I was doing this was in the green zone (i.e not the > production day). > > Every server had memory errors, somewhere, regardless of memory, etc. So I > pulled out of storage a brand new, never used Dell 1750 motherboard and put > a pair of brand new, never used 1gb sticks (from Dell, in the Dell > packaging) on that, ran memtest86, and it failed too. > > ::headscratch:: > > I booted every PC in the shop with memtest86, and every one of them failed > the test also. Now I'm suspect of that testing program. I shared all this > because I thought maybe you'd find this humorous. I did until I had to > reassemble everything back the way was at 2am :) > > > Dave Williams: > >>I would be interested to see what the kernel actually thinks the > >>processor/processors is/are and what processor specific options > >>are being loaded. Just a random thought. > > Fair question, but it seems Centos believes the processors to be Genuine > Intel. When I listed it out it showed four processors (two per core I > imagine) but I only pasted one here to keep the message somewhat short. > The > result was a surprise because one of my techs who is no longer with us had > upgraded this machine to 3.02Ghz processors. Seems they didn't make it > into > the machine !?!? > > cat /proc/cpuinfo > > processor : 3 > vendor_id : GenuineIntel > cpu family : 15 > model : 2 > model name : Intel(R) Xeon(TM) CPU 2.40GHz > stepping : 9 > cpu MHz : 2385.596 > cache size : 512 KB > physical id : 3 > siblings : 2 > core id : 0 > cpu cores : 1 > apicid : 7 > fdiv_bug : no > hlt_bug : no > f00f_bug : no > coma_bug : no > fpu : yes > fpu_exception : yes > cpuid level : 2 > wp : yes > flags : fpu vme de pse tsc msr pae mce cx8 apic mtrr pge mca cmov > pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe cid xtpr > bogomips : 4770.10 > > > --- > > If I goggle for "hlds_run line 321" I get no shortage of hits, and it seems > a lot of folks are having this problem in recent months with not to many > solutions. I wonder if steam/valve know about this or if those of us with > this issue just have unhappy hardware/OS. I found the linux engine 53 > binaries and copied them over, and got the same results, thinking going > back > one version might help. > > I tried this on another Dell server, and got the same results. Same if I > force the amd, i484 or i686 binaries to load, either server. > > This morning I pulled a Compaq DL580 out of the "to scrap" pile and ran the > same version of memtest86 as above, and the machine passed 100% despite > it's > vintage. Hooray Compaq. I get Linux on it this afternoon and continue > testing. > > The sad thing is I used to run several hlds-linux servers when day of > defeat > first came out as a mod to go over halflife, and wasn't a steam game. You > know, the old won-id network and all that. > > It's too bad steam keeps their software in binary form, a lot of problems > can be resolved by compiling server daemons on the server it is to run on. > > > _______________________________________________ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlds_linux > _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
