New Separation Runs

ps_separation_82_3s_dr8_3
de_separation_82_3s_dr8_3


These have increased the number of steps from

ps_separation_82_3s_dr8_2
de_separation_82_3s_dr8_2

Which should fix the errors people were reporting.

Please report problems here.

Thank you,
Jeff Thompson

Still failing with GPU (units 330082567, 330083256, 330082966). Someone completed 330082567 though.

The new Version: ps_separation_82-3s-dr8-3_1358941502_29605592
works also not,

http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=330033046 old one
http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=330079218 new one

Windows 7 Professionel
Radeon Sapphire 7950 3gb 810 Cpu MHZ
Driver Catalyst 13.1

Normal Mw Wu ps_separation_23_3s_sSgr_1_1358941502_29575819 works fine

The WU´s failling 0.00 seconds to maximum 0.05 seconds. cpu time
all 5 new wu`s are failling.

http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=429564623
http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=429563200
http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=429562269
http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=429561997
http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=429560857
http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=429559136
http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=429558309

All Wu´s are failling both of the new ones ps_separation and de_separation

de_separation_82_3s_dr8_2_1358941502_29558890_2 is reporting that it will take 87600:00:00 to complete on ATI graphics !!!!!!

System is not sending completed or calling for new ones at the present time. Unknown why other than it probably thinks I have plenty to do with that huge wu. *shrug*

---

I got the same time estimate, 87600:00:00, on an AMD card. The task locked up the BOINC client and it would not restart. I ended up renaming the milky way directory and restarting. A new directory was created and all the nbody's ended up in a computation error, I had suspended the GPU tasks prior to this so they were recovered and all is running again.

---

I will pull them back down and check through all the parameters. I will also try some different variants on the test machine before reposting..

Thank you for the feedback.


Jeff Thompson

I thought the nbody's were CPU only, the ones with the 87600:00:00 time estimate come in marked for the GPU. My system appears to try to start them on the GPU and I am wondering if this is what caused the crash?

Separation doesn't use the nbody application and has GPU support.


Jeff Thompson

I have 2 more on another system that is CPU processing only

de_separation_82_3s_dr8_1358941502_29567987_1
de_separation_82_3s_dr8_1358941502_29616483_0

Both showing the same time for completion 87600:00:00

Previous system running with ATI card has updated and now sending and receiving WUs. Don't know what happened to the one that had the extended time frame unless it borked and got tossed.

Edit - These WUs are now showing up as invalid with too many errors. Guess that sorts it out. 3 different systems processed them with same results.

---

I have several that reported 87600 hours to complete (curiously 3650 days) that I aborted as they forced others to high priority.

Computer 479613

de_separation_82-3s-dr8-3_1358941502_29610520
ps_separation_82-3s-dr8-3_1358941502_29611872
ps_separation_82-3s-dr8-3_1358941502_29611626
ps_separation_82-3s-dr8-3_1358941502_29614009
de_separation_82-3s-dr8-3_1358941502_29612832
ps_separation_82-3s-dr8-3_1358941502_29615173
ps_separation_82-3s-dr8-3_1358941502_29617289
de_separation_82-3s-dr8-3_1358941502_29613679

If you want to have a look see.

Well I am not sure what is happening but I woke up with the system down and unable to restart again. Again had to dump the Milkyway directory to recover.
The task was from the de_separation_82_3s_dr8_3 group.

Got two of these in row after a restart, which results in a client lockup. These always have the 87600:00:00 time estimate and always crash my client. This is making extremely difficult to run tasks on my system. Any chance these are going to be corrected?

ps_separation_82_3s_dr8_2_1358941502_29566368_1

<core_client_version>7.0.33</core_client_version>
<![CDATA[
<message>
- exit code -1073740940 (0xc0000374)
</message>
<stderr_txt>
BOINC: parse gpu_opencl_dev_index 0
<search_application> milkyway_separation 1.02 Windows x86 double OpenCL </search_application>
Unrecognized XML in project preferences: max_gfx_cpu_pct
Skipping: 5
Skipping: /max_gfx_cpu_pct
Unrecognized XML in project preferences: allow_non_preferred_apps
Skipping: 1
Skipping: /allow_non_preferred_apps
Unrecognized XML in project preferences: nbody_graphics_poll_period
Skipping: 30
Skipping: /nbody_graphics_poll_period
Unrecognized XML in project preferences: nbody_graphics_float_speed
Skipping: 5
Skipping: /nbody_graphics_float_speed
Unrecognized XML in project preferences: nbody_graphics_textured_point_size
Skipping: 250
Skipping: /nbody_graphics_textured_point_size
Unrecognized XML in project preferences: nbody_graphics_point_point_size
Skipping: 40
Skipping: /nbody_graphics_point_point_size
BOINC GPU type suggests using OpenCL vendor 'NVIDIA Corporation'
Error loading Lua script 'astronomy_parameters.txt': [string "number_parameters: 4..."]:1: '<name>' expected near '4'
Error reading astronomy parameters from file 'astronomy_parameters.txt'
Trying old parameters file
Integral area { 1819239215, 1769234796, 1008738314 } will overflow progress calculation

</stderr_txt>
]]>

---

I have several tasks with the same error.
de_separation_82-3s-dr8-3-test_1358941502_29860874_1
stderr output

<core_client_version>6.10.58</core_client_version>
<![CDATA[
<message>
process exited with code 1 (0x1, -255)
</message>
<stderr_txt>
<search_application> milkyway_separation 1.00 Linux x86_64 double </search_application>
Unrecognized XML in project preferences: max_gfx_cpu_pct
Skipping: 20
Skipping: /max_gfx_cpu_pct
Unrecognized XML in project preferences: apps_selected
Skipping: app_id
Skipping: app_id
Skipping: /apps_selected
Unrecognized XML in project preferences: nbody_graphics_poll_period
Skipping: 30
Skipping: /nbody_graphics_poll_period
Unrecognized XML in project preferences: nbody_graphics_float_speed
Skipping: 5
Skipping: /nbody_graphics_float_speed
Unrecognized XML in project preferences: nbody_graphics_textured_point_size
Skipping: 250
Skipping: /nbody_graphics_textured_point_size
Unrecognized XML in project preferences: nbody_graphics_point_point_size
Skipping: 40
Skipping: /nbody_graphics_point_point_size
Error loading Lua script 'astronomy_parameters.txt': [string "number_parameters: 4..."]:1: '<name>' expected near '4' 
Error reading astronomy parameters from file 'astronomy_parameters.txt'
  Trying old parameters file
Integral area { 1819239215, 1769234796, 1008738314 } will overflow progress calculation
Error reading parameters file
Failed to read parameters file
18:28:56 (17086): called boinc_finish

</stderr_txt>
]]>

Today, all the Milky Way runs have ended with a "Computational Error".

I received 1 download that would require 28,000+ hours to run. I aborted this run.

ps_separation_82 are still being released and they all crash within a second.

Once we close the old runs there are still some runs in the systems that process through.


I am using a versioning number at the end of the runs and that tells you which release it was.

de_p_82_3s_dr8_2
ps_p_82_3s_dr8_2

de_p_82_3s_dr8_3
ps_p_82_3s_dr8_3

Were the previous runs I have re-confirmed that they are not active runs. So old units were still in the queue with the other separation runs.

I have

de_p_82_3s_dr8_4
ps_p_82_3s_dr8_4

starting and am posting on the changes on those in the next thread.



Jeff Thompson