So will the double credit days start today/night @00:00:01 ?

I do not know for other users here, but i have several computers running 24/7 for Milkyway and since the last update, my average credit total drop from a usual 250000 to the 150000 ... I even added a brand new three core AMD computer to my pool last week, so I find it very odd. It seems to back log a little more since i have more then 200 tasks completed and waiting validation. But it seems that one WU out of four gets the invalid flag.

As for the double credits question, since I crunch 24/7, anytime will do, but I do prefer during the week, since my gaming pc go back in the crunching pool during that time ;p

Anyhow, cheers to everyone here, especially the volunteers for their great work and project :)

I do not know for other users here, but i have several computers running 24/7 for Milkyway and since the last update, my average credit total drop from a usual 250000 to the 150000 ... I even added a brand new three core AMD computer to my pool last week, so I find it very odd. It seems to back log a little more since i have more then 200 tasks completed and waiting validation. But it seems that one WU out of four gets the invalid flag.

I have had such problems with the BOINC-Client 6.12.26. I had a lot of invalid WU, but after changing to an other version (now using 6.12.28) the problem disappeared. What version do you use?

---

A random start to double credit probably would be fairer for us 24/7 long term crunchers, as I'm quite sure that there are a multitude of credit hounds out there, just waiting to log in the moment they hear it's started, and who will disappear shortly afterwards.

But Northern Europe and other places are having a mini heatwave this week, and some kit will be shut down, so I would say go now, or leave it until Thursday/Friday.

---

Don't drink water, that's the stuff that rusts pipes

Dont know if it is possible to do sooo much Credits on two days für credit hounds (i know some millions are possible, but gives primegrid and collaz not more credits for that hounds?). But well, they do work for science when use the electricity for MW shortly :) Normaly i only crunch MW @ weekend for some hours, and MW was often down on Weekend, so i got often nothing to do for the "whole week". (energy too expensive here because in summer i have to run the AC too then) When someone wants to randomies it, then tell the week and i run the machine for the full week, but please not more "randomes" over a month or so. omg my english makes me angry sometimes :D

---

DSKAG Austria Research Team: http://www.research.dskag.at

I have had such problems with the BOINC-Client 6.12.26. I had a lot of invalid WU, but after changing to an other version (now using 6.12.28) the problem disappeared. What version do you use?.

I do have the older client version installed on my computers, i will give the new one a try tonight. Thanks for the tip.

This is almost what I sugguested,but with randomly I mean not telling when the double credit starts. The dedicated MW crunchers will see it. I am a dedicated night and weekend cruncher as the pc's with high end GPU cards are all 800-1000W machines (they don't use so much energy though). I have only a light pc running 24/7 and that is doing Einstein@home (and very occasionally LHC).
By the way the forecasted 33°C for today at my place (eastern Netherlands) was way to high (29.6°C at the moment). So no minor heatwave. I will crunch overnight.

---

Greetings from,
TJ

Sure seem to be an unusual number of "validate errors" today on most machines...

So I'm out of town for the remainer of this week. So I think we'll be doing them next week. It looks like there couple more bugs we need to fix so everyone gets their fair share of the double credit.

---

Looks like you're practically out of workunits for now anyway.

Should I take that as there will be no new wu's until next week then?

Should I take that as there will be no new wu's until next week then?

There were just some hard drive problems last night, and I ended up passing out while waiting for them to become available again. Now that I'm somewhat awake, I fixed the problems that were happening last night and everything should be running smoothly again.

---

Should I take that as there will be no new wu's until next week then?

There were just some hard drive problems last night, and I ended up passing out while waiting for them to become available again. Now that I'm somewhat awake, I fixed the problems that were happening last night and everything should be running smoothly again.

Hmm... maybe it will but it does not look like that.
'Server Status' page report "Tasks ready to send 1"

EDIT: woops! just changed to "Tasks ready to send 7,471"

I'm getting major computation errors on two systems. System-1 has (2) hd 4850 running Win7 x64 and System-2 has (1) HD 4850 running WinXP x64.

Here are the bad tasks

Task___Work unit___Computer___Run time___CPU time

60844670 --- 41501809 --- 292090 --- 276.03 --- 2.39
60840401 --- 41499775 --- 292090 --- 002.03 --- 0.00
60955989 --- 41534254 --- 293129 --- 327.22 --- 3.74
60856587 --- 41500158 --- 292090 --- 002.02 --- 0.02
60844700 --- 41501839 --- 292090 --- 275.44 --- 2.38
60844697 --- 41501836 --- 292090 --- 275.66 --- 2.42
60844696 --- 41501835 --- 292090 --- 276.36 --- 2.45
60844695 --- 41501834 --- 292090 --- 277.05 --- 2.50
60970392 --- 41541105 --- 293129 --- 323.75 --- 3.32
60970391 --- 41539593 --- 293129 --- 325.32 --- 3.96
60970388 --- 41516742 --- 293129 --- 323.59 --- 3.78
60970364 --- 41501635 --- 293129 --- 272.63 --- 2.12
60970363 --- 41501209 --- 293129 --- 270.98 --- 1.81
60970396 --- 41544563 --- 293129 --- 328.11 --- 3.70
60970395 --- 41544474 --- 293129 --- 325.05 --- 3.87
60970394 --- 41542684 --- 293129 --- 323.57 --- 3.68
60970393 --- 41541244 --- 293129 --- 325.39 --- 3.62

Stderr output example

<core_client_version>6.10.58</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
Using SSE2 path
Found 1 CAL devices
Chose device 0

Device target: CAL_TARGET_770
Revision: 2
CAL Version: 1.4.838
Engine clock: 700 Mhz
Memory clock: 850 Mhz
GPU RAM: 512
Wavefront size: 64
Double precision: CAL_TRUE
Compute shader: CAL_TRUE
Number SIMD: 10
Number shader engines: 1
Pitch alignment: 256
Surface alignment: 256
Max size 2D: { 8192, 8192 }

Estimated iteration time 354.087500 ms
Target frequency 30.000000 Hz, polling mode 1
Dividing into 10 chunks, initially sleeping for 0 ms
Integration range: { nu_steps = 640, mu_steps = 1600, r_steps = 1400 }
Using 10 chunk(s) with sizes: 160 160 160 160 160 160 160 160 160 160
Integration time = 217.939909 s, average per iteration = 340.531108 ms
Integral 0 time = 219.282973 s
Estimated iteration time 88.521875 ms
Target frequency 30.000000 Hz, polling mode 1
Dividing into 2 chunks, initially sleeping for 0 ms
Integration range: { nu_steps = 640, mu_steps = 400, r_steps = 1400 }
Using 2 chunk(s) with sizes: 192 208
Integration time = 53.570695 s, average per iteration = 83.704212 ms
Integral 1 time = 53.878763 s
Estimated iteration time 309.826563 ms
Target frequency 30.000000 Hz, polling mode 1
Dividing into 9 chunks, initially sleeping for 0 ms
Integration range: { nu_steps = 640, mu_steps = 1400, r_steps = 1400 }
Using 9 chunk(s) with sizes: 144 160 160 144 160 160 160 144 160
Chunk mu steps does not match: 1392 != 1400
Integral 2 time = 0.797812 s
Failed to calculate integral 2
02:52:10 (2932): called boinc_finish

</stderr_txt>
]]>

Any ideas????

---

"Those who would give up essential Liberty, to purchase a little temporary Safety, deserve neither Liberty nor Safety." (Benjamin Franklin)

Same here, win 7 64 bits and win xp 32 bit. Ati 5850.

They are doing a lot of test runs right now, see this thread:
http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=2489

Not looked at all errors you posted but those I did were all the test WUs. At present the test WUs are going full term and erroring out in the last 2 seconds consolidation.

At present it appears its this bit going wrong:

Using 9 chunk(s) with sizes: 144 160 160 144 160 160 160 144 160
Chunk mu steps does not match: 1392 != 1400
Integral 2 time = 0.797812 s
Failed to calculate integral 2
02:52:10 (2932): called boinc_finish

Most ps 10's and 13's WUs are working for most people, so its worth running those to see if they work for you rather than abort them just because of the test apps. We await developments tomorrow.

Regards
Zy

I've had 1 test wu that failed today.

Name ps_separation_17_test2_6251_1
Workunit 41539444
Created 29 Jun 2011 | 0:09:36 UTC
Sent 29 Jun 2011 | 0:17:26 UTC
Received 29 Jun 2011 | 2:21:51 UTC
Server state Over
Outcome Computation error
Client state Compute error
Exit status -1073740940 (0xffffffffc0000374)
Computer ID 293167
Report deadline 11 Jul 2011 | 0:17:26 UTC
Run time 897.95
CPU time 436.04
Validate state Invalid
Credit 0.00
Application version MilkyWay@Home v0.82 (cuda_opencl)

Two others came up inconclusive: wu 40332765 and wu 41529440.
i don't think this is the same cause as the test2 wu, but decided to post for your review in case it would be helpful.

Just had another test2 wu fail in the last few seconds. computational error.

Name ps_separation_17_test2_7179_3
Workunit 41544572
Created 29 Jun 2011 | 0:32:27 UTC
Sent 29 Jun 2011 | 0:34:16 UTC
Received 29 Jun 2011 | 2:44:16 UTC
Server state Over
Outcome Computation error
Client state Compute error
Exit status -1073740940 (0xffffffffc0000374)
Computer ID 293167
Report deadline 11 Jul 2011 | 0:34:16 UTC
Run time 858.38
CPU time 365.71
Validate state Invalid
Credit 0.00
Application version MilkyWay@Home v0.82 (cuda_opencl)

should i delete any other test2 wu's or is it more helpful for me to let them run and report them?

Getting some 0.82 (cuda_opencl) computation errors...

Is this project back up to snuff again? Can I unsuspend it now?

---

32bit Windows XP Home
AMD Opteron 180
ASUS A8N-SLI Motherboard
Nvidia 450GTS GPU
4GB DDR Memory