DNETC and MilkyWay down

Toggle sidebar Toggle sidebar
petrusbroder

petrusbroder

Elite Member
Nov 28, 2004
13,348
1,155
126
I noticed that my comps have an unusual number of uploads and "ready-to-send" results ... and realised that DNETC and MilkyWay are both down.

I can't get a server status either because their web-sites are down too.


We are talking bout 450 WUs in DNETC and 250 in MilkyWay - and I will be running out of WUs from those projects too ...

Edit:
MilkyWay problem solved: a super crash of the server with WUs and results lost. Servers are up and WUs are rolling out.
DNETC problem of jus´dicial nature - will take some time to solve. It appears that the server was seized by the police suspected for illegal file sharing. But therse are just rumors as yet.
 
Last edited:
Sunny129

Sunny129

Diamond Member
Nov 14, 2000
4,823
6
81
yeah Milkyway has been down for about 48 hours now. i got 3 intermittent WU's yesterday afternoon, but then the servers went silent again. i've switched my 5870 GPU over to S@H MB & AP tasks for the time being...
 
petrusbroder

petrusbroder

Elite Member
Nov 28, 2004
13,348
1,155
126
Yupp. I lost crunched 1300+ WUs, plus a much bigger numer un-crunched WUs ...
Well, as sombody on the MW-forums said: "It is only a number of points and a few cents of electricity" 4 days of GPU-crunching on three GPUs with a GPU taking some 150W/h = 24h*4*3*150W/h = 43 200Wh = 43.2kWh = approx US$ 10. Not too much, but I still really dislike that cost ... Oh, BTW, the CPUs crunched other projects ...
 
petrusbroder

petrusbroder

Elite Member
Nov 28, 2004
13,348
1,155
126
Well, yes and no. There are no WUs as of 10 minutes ago ... but the other pages are up.
 
Sunny129

Sunny129

Diamond Member
Nov 14, 2000
4,823
6
81
sending out some more workunits
I'm trying something new. I'm setting the a max total, max CPU, and max GPU workunits in the config_aux.xml file. Maybe this will put a cap on things.

Right now the total max will be 16, the cpu max (per processor) will be 3, and the gpu max (per processor) will also be 3. If this seems to work, and the server can handle it we'll start to increase these limits and see how things go. 10 Apr 2011 | 1:04:51 UTC · Comment
Click to expand...

i've been getting a steady flow of new WU's, all of which have been validating and earning credit so far. note that there is an updated app, v0.57. i don't know if it'll automatically download and update the v0.23 .exe file in the MW@H project folder, or if you manually have to detach from the project and reattach to get v0.57. if anyone has problems or gets errors, check out the following threads on the MW@H message boards:

ATI application (v0.57) should be available now

sending out some more workunits


happy crunching...
 
petrusbroder

petrusbroder

Elite Member
Nov 28, 2004
13,348
1,155
126
Thanks for the info!
I have new WUs. as far as I can see they are crunching well.

MilkyWay-problem solved ...

DNETC will probably not come up until the judiciary issues are resolved.
 
zzuupp

zzuupp

Lifer
Jul 6, 2008
14,866
2,319
126
v0.57 downloaded automatically. I'm getting new WUs. :)
These seem to be taking longer, and using a lot more CPU time
 
Sunny129

Sunny129

Diamond Member
Nov 14, 2000
4,823
6
81
zzuupp said:
v0.57 downloaded automatically. I'm getting new WUs. :)
These seem to be taking longer, and using a lot more CPU time
Click to expand...
you should post the details of your situation in one of the two MW@H threads i linked above. the developers and testers are trying to get as much info as possible in order eliminate folks' problems...which seem to stem from various combinations of OS (32-bit vs 64-bit), BOINC version, GPU, GPU driver version, etc. the issues you're experiencing are quite common over there, so you're not alone in that respect. the new MW@H v0.57 app seems a bit finicky right now, but Matt and Travis have been working hard to isolate and eliminate the bugs.

luckily i'm crunching away error-free. here are the specifics of my platform:
- Windows XP Pro SP3 32-bit
- HD 5870 2GB GPU
- Catalyst 11.3 GPU drivers
- BOINC v6.12.18

my crunch times have not increased, nor has my CPU usage - its still the standard 5% of a CPU core per GPU task (0.05 CPU's + 1.00 ATI GPUs).
 
Sunny129

Sunny129

Diamond Member
Nov 14, 2000
4,823
6
81
for those of you who aren't following the action closely on the MW@H message boards, be sure to check in at least once a day. the powers that be are planning on somehow compensating everyone credit-wise for all the down time we experienced over the weekend. the talk is that it'll come in the form of a 1 or 2-day long period in which all points earned will be worth double the credit. or perhaps it might just be an automatic points award...i don't know. but assuming they can get all the bugs worked out in the next few days, i'm sure the compensation period will immediately follow. so stay tuned...
 
Sunny129

Sunny129

Diamond Member
Nov 14, 2000
4,823
6
81
Assimilator1 said:
I don't have time to go there as well as here ;), what's the latest mate?
Click to expand...
1) well things have been running pretty smoothly for most folks now that the MW@H app is up to version 0.62. specifically, its fixed alot of lag problems that people with a wide variety of hardware combinations have been experiencing (or at least the programmer has given us some parameters to play with that can help with GUI lag).

2) also, GPU utilization seems to be up...on some of the previous fixes, people were getting GPU usage in the 60-85% range (and possibly outside that range as well). with v0.62 it seems like more people are seeing 90-100% GPU utilization.

3) the other big issue addressed was the amount of CPU usage required to run GPU tasks. before, some folks were losing an entire CPU core to MW@H GPU tasks, and in some cases more than one CPU core was getting used up by MW@H GPU tasks. with v0.62 it seems as though alot more folks are experiencing a normal 5% CPU load per GPU task (0.05 CPUs + 1.00 ATI GPUs).

although the 2nd and 3rd issues are treated independently in many instances, they are sometimes linked. for instance, when i would try to run Einstein@Home on all 6 CPU cores while a MW@H GPU task was running, it would drop my GPU load down to ~85 on average, thus slightly increasing my GPU task run times. if i would change the "on multiprocessor systems, use at most __% of the processors" to 85% and consequently reduce the number of CPU cores crunching E@H from 6 to 5 (thus freeing up a core to handle any MW@H GPU task instructions), my GPU load would go back up to 97-100%. with the new MW@H v0.62, i can run E@H on all 6 CPU cores without a GPU utilization penalty or an increase in GPU task run times.

as far as the "double points" bonus goes, still no word on it. i'm not surprised that its been pushed back this much. they've had alot of bugs to fix.
 
Assimilator1

Assimilator1

Elite Member
Nov 4, 1999
24,165
524
126
Thanks for the update Sunny :), so at last their's a way to allievate GUI lag?, sweet :cool:.

I need to try that as with the latest update the lag is so bad I can't use the pc at all & just shutting off MW had become nearly impossible,today I used the task manager & ended task! (where's as before I could use my pc as long as I didn't have lots of window switching & scrolling to do;)),

I'll have to check issues 2&3, not done so recently.
 
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom