TPU's WCG/BOINC Team

[TRWOV]

how do I get the logs?

edit:
Started with the live CD and tried to access the disk on the file manager. This is the error message:

Error mounting /dev/sda1 at /media/lubuntu/963036e6-f5ed-4825-d3c1749c06ff: Command-line `mount -t "ext4" -o "uhelper=udisks2,nodev,nosuid" "/dev/sda1" "/media/lubuntu/963036e6-f5ed-4825-d3c1749c06ff"' exited with non-zero exit status 32: no medium found on /dev/sda1


edit2: plugged the disk into my Win8 machine and shows up fine in compmgmt.msc so it's not dead

 

[[Ion]]

This is new. The Lubuntu cruncher showed a blank screen with a cursor on the top. A hard reset brings up a GNU GRUB menu with the following options:

-Ubuntu
-Ubuntu advanced options
-+Ubuntu , kernel 3.whatever-generic
-+Ubuntu, kernel 3.whatever-generic (recovery mode)
-Memory test (memtest86+)
-Memory test (memtest86+, serial console 11whatever)

Every option to launch the OS takes me to a blank screen. I waited about 10 min and nothing showed up.

Memtest reports no problems (everything is running at stock settings so no surprises here).



Gonna test with Mint now.

The general difficulties that Linux gives are the reason that many of us choose Windows, even with the lower output it brings.

 

[TRWOV]

I'm reading that ext4 has "rare instabilities" (maybe my super power is bringing that out). Is that true? Should I try ext3?

 

[Jstn7477]

I don't understand why your Linux installs only work once and then "die." I've had less than 1yo Ubuntu distros installed on everything ranging from a 2002 Sony P4 desktop to a Haswell i7, and none of my installs have done that ever, even when the drive was pulled and inserted into different machines. The only rig I couldn't get it to install on was a Biostar A780L3B 760G board that would freeze while I was entering my user information. Thing is, 90% of the systems I have installed it on are Intel with NVIDIA or Intel GPUs excluding my main desktop which is Intel with an HD 7970 or a 2005 Turion64 laptop with ATI XPress 200M graphics.

Do you attempt to install video drivers when you set these machines up, or do you just install updates and BOINC and that's it? I simply cannot understand what your machines are doing. If there is any sort of video issue, Linux should boot into low graphics mode or just a straight up terminal asking for your login if it's that bad.

 

[TRWOV]

Well, the first time I installed Ubuntu on a USB drive. It ran fine for a few days, just having some problems with stuck WUs from time to time but one day it was just locked up and after a hard reset it wouldn't load again. I thought it was because of the USB stick.

Second try was on a 8GB HDD (leftover from a Xbox HDD upgrade). This time I deleted some stuff since after installation there where about 1.2GB free on the drive and I wanted at least 2GB to cover for CEP2 requirements. Somehow I deleted the desktop enviroment and had to sudo it back out. After restoring it I installed boinc and it ran for some days. I came back from work one day and it was frozen again.

About two or three weeks ago I tried again when the linux vs windows ppd disparity was discussed. This time I did nothing but install Ubuntu then the boinc manager, setting it up and never touching it again. To be fair, this install was done on a cruncher with a 1055t OCed to 3.5Ghz (which was fully stable in Windows) so I don't know if that had a hand on it. It survided for a couple of days.

Today I set a new record, not even 24 hours have passed. Earlier today I lost a VNC connection to the cruncher and when I got home from work BAM! blank screen.

 

[Chicken Patty]

had to shut WCG/F@H down for a bit, its just too hot in my apartment with the heat lately and ive been getting high 70s temp wise and that doesnt sit well with me

Same here about a night ago. Went out for my birthday with my girlfriend straight from work. House had the AC off and all three rigs going for pretty much well over 12 hours. Got back home to sleep and I just couldn't take it. Turned them back on the next morning.

 

[TRWOV]

LOL... "Welcome screen is not responding"

This isn't a good start.

 

[Chicken Patty]

LOL... "Welcome screen is not responding"

This isn't a good start.

Geez, just doesn't wanna cooperate heh?

 

[TRWOV]

Installing boinc now. Apart from the Welcome screen fail Mint feels faster. Maybe it uses another file system?

edit: welcome screen and screen lockup disabled.

edit2: boinc installed and running. Where can I set Mint to automatically accept incoming RFB connections?

edit3: NVM, desktop sharing

 

[librin.so.1]

Maybe it uses another file system?

Was just about to suggest that. As it does seem like a filesystem failure, as I was suspecting.
Though, are You sure this old eksbawks HDD is not failing in the first place? As ext4 support is finished up / stable / "ironed out" for a long time now, I would find it really odd if things borked up because of it like this a couple of times in a row.
(BTW I use ext4 for more than a year now and never got a single issue with it)

 

[TRWOV]

It passed every Seagate Tools test last night so I'd say the disk is fine.

 

[ThE_MaD_ShOt]

TRWOV, man you do have some bad luck with Linux. My rig that is running it, the os has been through hell and back and shouldn't be running but it still ticks on. I originally installed it on my mini s939 3800+ rig. have since bounced that hdd with the same os into multiple rigs and besides the occasional upgrade here and there it soldiers on. I haven't installed a fresh copy of the os since 10. something.

 

[librin.so.1]

In that case I'd say try FAT32. I know FAT32 is quite a pile of horse poop, but it is a failsafe[1] filesystem.

[1]The OS-level implementation is as solid as it can get, thus any failures with it would simply point at hardware.

 

[Jstn7477]

TRWOV, man you do have some bad luck with Linux. My rig that is running it, the os has been through hell and back and shouldn't be running but it still ticks on. I originally installed it on my mini s939 3800+ rig. have since bounced that hdd with the same os into multiple rigs and besides the occasional upgrade here and there it soldiers on. I haven't installed a fresh copy of the os since 10. something.

Same, whenever I scrap or change a Linux rig around, I just pull out the drive and put it in my drawer to be placed into the next machine. Haven't had to install it on a machine since December 2012.

 

[librin.so.1]

My longest Linux installation laster over 4 years. Till I finally killed it off with my cruel and unusual testing and curiosity.

 

[TRWOV]

I don't know. Maybe it's the chipset? I've run linux in different configurations but the common factor is the SB710 chipset.




In other news I think my plan to stick with 8320s for the crunchers won't go. I've read that there's basically zero overcloking for FX-8xxx on this board (8320 get to 3.7Ghz and that's it) so if I want 4Ghz I should buy the 8350 outright. There's a new board revision out (5.0) so maybe (maybe) it would OC better than my current 4.1.

I guess I'll buy an 8320 and a rev.5 board to find out. Worst case scenario I can stick the 8320 on the CCTV PC at work. It would still offer better ppd than the X4 620 anyway.

 

[ThE_MaD_ShOt]

I don't know. Maybe it's the chipset? I've run linux in different configurations but the common factor is the SB710 chipset.

Not sure and I can't remember if I ran this drive on a sb710 or not. Right now it's on a sb750. I think I may have ran an install on a Msi board that had a sb710 but I am not 100% sure.



I do still have 1 board that runs a sb710. I will try and convince my wife she doesn't need to use her rig for a few days and I will test it out on it. I have a spare hdd I can use.

 

[librin.so.1]

Reading through the net a bit, it seems this chipset has issues due to faulty hardware design. For starters, disable HPET (High Precision Event Timer) in the BIOS, add a "hpet=disable" kernel boot option and make sure Your HT link is running at no less than 1.8 GHz. I believe that should make those issues go away.
HPET is of no use for things like BOINC, so disabling it won't have a negative impact.

 

[theonedub]

Have had my PC off for the past week, 7 straight days at an average temp of 106F it was hot. Will be firing it back up this weekend.

 

[TRWOV]

I need a hug...

kernel panic - not syncing: Attemping to kill init! exitcode=0x00000600

drm_kms_helper: panic ocurred, switching back to text console

I'll disable HPET and try again. (-_-)

 

[[Ion]]

Have had my PC off for the past week, 7 straight days at an average temp of 106F it was hot. Will be firing it back up this weekend.

Awesome! Let's do this thing!

kernel panic - not syncing: Attemping to kill init! exitcode=0x00000600

drm_kms_helper: panic ocurred, switching back to text console

I'll disable HPET and try again. (-_-)

I believe that sort of an error is sub-optimal...

 

[TRWOV]

I guess I'll setup a USB stick with a persistence file. Maybe that'll work.

 

[[Ion]]

I guess I'll setup a USB stick with a persistence file. Maybe that'll work.

That has worked well for me in the past. You seem to be having rotten luck though, so I'm not sure...

 

[Peter1986C]

kernel panic - not syncing: Attemping to kill init! exitcode=0x00000600

drm_kms_helper: panic ocurred, switching back to text console

I'll disable HPET and try again. (-_-)

So the graphics driver shits his pants? What graphics hardware are you using, and did you try a propietary driver (if any are available, incase of AMD and nvidia)? You seem to have been using the open source Radeon/Nouveau/Intel graphics drivers. Just in case, check in the BIOS whether the PCIe Link Speed is set at 100Mhz and not even a Hz higher (nor lower). I have had a graphics board act up severely because of wrong link speeds.

 

[TRWOV]

I'm using whatever Mint had preloaded/downloaded. Graphics are HD3200 IGP.

Installed again with HPET disabled. boinc is working right now and so far has completed 8 tasks. Normally I'd report them manually but I don't want to disturb this install at the moment.