Due to the validation issues, I've updated the CPU applications. Hopefully this should fix the problem. Let me know how they're working here.



--Travis

---

The PPC Mac applications appear to have been removed:

Thu  7 Oct 21:51:11 2010	Milkyway@home	Message from server: No work sent
Thu  7 Oct 21:51:11 2010	Milkyway@home	Message from server: MilkyWay@Home N-Body Simulation is not available for your type of computer.
Thu  7 Oct 21:51:11 2010	Milkyway@home	Message from server: MilkyWay@Home is not available for your type of computer.

---

Version 0.40 CPU app seems to be behaving itself run-time-wise under Windows. (None returned from here yet for validation though.)

However, WU 162332342 for "de_12_3s_5_76295_1286490124" identifies itself as version "0.04". Is this intentional? The others show "0.40".

EDIT: I detached and reattached for good measure. MW says it is downloading app "milkyway_0.4_windows_intelx86.exe". But CPU WU 162623844 "de_11_2s_5_363797_1286532448" it downloaded immediately thereafter shows an application name of "MilkyWay@Home 0.04".

hackgrav: tree-incest detected

I got an error with this message in the log, see the unit log here
http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=213678672

I'm using milkyway_nbody_0.07_windows_intelx86__sse2.exe

What's this tree-incest?

For whatever it's worth to you. The new version 0.40 (milkyway_0.40_windows_x86_64.exe) is running on my Windows machine, but much slower compared to the version 0.19. Where the 0.19 processes 0.001-0.002% per second (I'm a small cruncher), the 0.40 version processes 0.001% per 2 or 3 seconds.
The memory usage of this new version is much lower too (0.19 is on 3100kB and 0.40 on 365kB).
Might be caused by the WU, I don't know(http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=162711434).

Jan

For whatever it's worth to you. The new version 0.40 (milkyway_0.40_windows_x86_64.exe) is running on my Windows machine, but much slower compared to the version 0.19.

Just checked my 0.40 WU's. All of them are going to run at least 50% longer than 0.19 typically did. One of them is going to run for a whopping 31hrs on a XEON E5540.

What happened?

After almost 3:25 hours of processing this WU 162711434 the estimated remaining time increased with 3:55 hours and it's not even on 5% completion!?!
I see Brian has a similar problem. I'm running on an i7 Q720 (on the CPU, my GPU cannot handle double precision).
If I can choose gives us back the 0.19 version.

Yup, I'm still getting the same message on my PPC's

Anything I can do to get back to work?


Fri Oct 8 13:10:58 2010 Milkyway@home Message from server: No work sent
Fri Oct 8 13:10:58 2010 Milkyway@home Message from server: MilkyWay@Home N-Body Simulation is not available for your type of computer.
Fri Oct 8 13:10:58 2010 Milkyway@home Message from server: MilkyWay@Home is not available for your type of computer.

After 5hr just over 7% completed, meaning this WU would take 60hrs where 18-19 is normal. I stopped this WU (and all others of v0.40). Temporarily no more Milkyway for me...

The Windows 0.4 binary is mysteriously 2x slower than the same thing on Linux/OS X. We're investigating why.

I haven't received any ATI WU's in the last few days, but the Server Status shows 100% OK. I've got 2 5850's sitting idle...

@RadonPL

If you are using BOINC Client 6.11.9, switch back to 6.11.7
6.11.9 contains an error that prevents users from downloading ATI WUs for any project.

---

When life gives you lemons, make lemonade!

What is really painful is to watch one of the new WUs process, then get to that 20-22 hour mark and end in what looks like a 'normal' end of processing, but show up as a computation error. So far, this is about 6 cpu-days of processing that is thrown away. I am close to aborting all MW in the queue and going elsewhere for a while.

Something is frakked.

http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=215114424

:(

Same problems. 0.04 application, a strange RAC peak some days ago then only validation errors and always "0.02 claimed credits".

Grrr!

Have I to detach and re-install, or is this a project issue as it seems?

The Dreamer

---

Just updated to Boinc 6.10.58 and still getting


Sat Oct 9 18:19:34 2010 Milkyway@home Message from server: No work sent
Sat Oct 9 18:19:34 2010 Milkyway@home Message from server: MilkyWay@Home N-Body Simulation is not available for your type of computer.
Sat Oct 9 18:19:34 2010 Milkyway@home Message from server: MilkyWay@Home is not available for your type of computer.

What is really painful is to watch one of the new WUs process, then get to that 20-22 hour mark and end in what looks like a 'normal' end of processing, but show up as a computation error.

Agree. Up to 37 hours here on fast XEON's and it croaks. All of the version 0.40/0.04 are dying on file xfer code -161. Will not be processing any more CPU WU's until this is fixed.

Another PPC user here not able to work on this project as of today. I'm wondering if this is intentional (PPC is kinda old...) or simply an oversight. I'll leave it on my BOINC for now, but if it doesn't get anything in a week or two, I'll just assume support was dropped and detach.

What is really painful is to watch one of the new WUs process, then get to that 20-22 hour mark and end in what looks like a 'normal' end of processing, but show up as a computation error.

Agree. Up to 37 hours here on fast XEON's and it croaks. All of the version 0.40/0.04 are dying on file xfer code -161. Will not be processing any more CPU WU's until this is fixed.

Is this for the nbody application or the other one?

---

Is this for the nbody application or the other one?

None of the WU's had "nbody" anywhere in the name to my recollection. The executable it was using: "milkyway_0.4_windows_intelx86.exe". The ones that errored out have all scrolled off my task list. All of the ones I looked at had names prefixed with DE and a two-digit number. I don't know why some show up as version 0.40 and others 0.04.

Check out CPU ID 161823 and 200321 in your database for more details.