Hello All,

We are happy to announce a new version of the Separation Modified Fit application. Version 1.32 will introduce a few new features including:

New Optimization Parameters
Support for Windows 32-bit
A Few Bug Fixes

Along with this release I would like to mention that I also started a couple of new runs:

de_modfit_15_3s_132_wrap_1
de_modfit_15_3s_132_wrap_2
de_modfit_16TestStars_1s_132_wrap_1
ps_modfit_15_3s_132_wrap_1
ps_modfit_16TestStars_1s_132_wrap_1

As per usual, if you notice any issues relating to these runs or this release version, please post them here and I will be glad to help.

Thank you for your continued support,
Jake W.

Im using boinc 7.4.12 x64
Most if not all of the TestStars ATI GPU WU's are showing up as 5GF in boinc and immediately error upon execution. This is happening for both de_modfit_16TestStars_1s_132_wrap_1 and ps_modfit_16TestStars_1s_132_wrap_1 type workunits.

System:
http://milkyway.cs.rpi.edu/milkyway/show_host_detail.php?hostid=585984

13 errors so far .. all of them on the ps_modfit_16TestStars_1s_132_wrap units. All of these erred within 2 seconds of starting

114 errors so far and counting. All the errors I am seeing concern "ps_modfit_16TestStars_1s_132_wrap_1" WUs. On the other hand, "de_modfit_15_3s_132_wrap_1" WUs are executing without any problem. I don't suppose there's anything I could do at my end to make the scheduler assign only the "de_modfit" and not the "ps_modfit" WUs to my machine? This problem is not good for the RACs. OK, OK, just asking.

---

I took those runs down. Seems to be an error with the parameter file. Will fix it and put new up later today. Thanks guys and sorry for the inconvenience.

Jake W.

Both the latest Modified Fit app and the N-Body app fail on every WU due to this error:
../../projects/milkyway.cs.rpi.edu_milkyway/milkyway_nbody_1.42_x86_64-pc-linux-gnu__mt: /lib64/libc.so.6: version `GLIBC_2.14' not found (required by ../../projects/milkyway.cs.rpi.edu_milkyway/milkyway_nbody_1.42_x86_64-pc-linux-gnu__mt)

Therefore, I can no longer run any Milkyway applications. :(

Sorry, but if you do not fix, I must detach from this project.

DJStarfox,

N-Body seems to be running fine on my Linux box.

Since you have your computers hidden, I can't see how they are configured. What version of Linux are you using? Is it a 64-bit version? If it is 64-bit, did you install the 32-bit compatibility libraries?

Have you ever been able to run N-Body and Modified Fit tasks successfully?

Hey DJStarfox,

We try to compile against older glibc libraries but we changed our build system for this release. I am looking into it now and we could possibly have a new release fixing the library compatibility issue soon. In the meantime you are welcome do compile the binaries on your own computer using the libraries you have. The code is on github https://github.com/Milkyway-at-home/milkywayathome_client. If you need help for this let me know and I will gladly PM you and give you help.


Additionally, could you post which version of glibc you currently use?

Sorry,

Jake W.

Thanks. I'm running Centos 6.5 x86_64 with glibc 2.12.

My developer skills are rusty at the moment, but I'll see if I can find time this weekend to compile the code.

Okay usually we compile on 2.7 and we should be releasing a new version of modified fit and nbody in a week or so compiled against these older libraries for Linux users. I will make another news post when this happens.

I took those runs down. Seems to be an error with the parameter file. Will fix it and put new up later today. Thanks guys and sorry for the inconvenience.

Jake W.

Even the new "fixed" WUs are ending in immediate errors

I haven't put new fixed runs up yet sorry. I got distracted working on the library issue. The work units you see giving errors are just the broken ones clearing out of the work unit queues on the server. They should all be over soon.

Jake

I had some workunits error out. I suppose the server queue still has some workunits that haven't reached their max error #...

ps_modfit_16TestStars_1s_132_wrap_2_1405680903_9442119_1
ps_modfit_16TestStars_1s_132_wrap_2_1405680903_9442124_1
de_modfit_16TestStars_1s_132_wrap_2_1405680903_9441944_1
ps_modfit_16TestStars_1s_132_wrap_2_1405680903_9442123_1
ps_modfit_16TestStars_1s_132_wrap_2_1405680903_9442197_1
ps_modfit_16TestStars_1s_132_wrap_2_1405680903_9442196_1
ps_modfit_16TestStars_1s_132_wrap_2_1405680903_9442122_1
de_modfit_16TestStars_1s_132_wrap_2_1405680903_9418202_1

bad astronomy_parameters.txt file.

Also... is the ver. 1.32 Nvidia OpenCL GPU app for Windows 64 bit the same as the ver. 1.30 app? There doesn't seem to be a ver 1.32 app for Windows 64 bit Nvidia.

I am having the same problem and on initial inspection al the failing units belong in this group:-

de_modfit_16TestStars_1s_132_wrap_1

At least I am not alone and do we know what the problem is?

Cheers

James

For what its worth, I'm still experiencing errors on the WUs de_modfit_16TestStars_1s_132_etc. When I first reported this on Sept. 9th, I thought you were fixing this.

---

I took down the run that was causing errors on Tuesday, those take a day or two to completely clear out of the workunit queue. However, yesterday I put a new version of the run up with what I thought was a fixed parameter file. Everything ran fine locally. I am currently working on taking the new run down which is also causing errors and fixing the parameter file.

Sorry,

Jake W.

Lack of required file on the server
required file milkyway_separation__modified_fit_1.32_windows_x86_64__opencl_amd_ati.exe

files available
milkyway_separation__modified_fit_1.32_windows_intelx86__opencl_amd_ati.exe
milkyway_separation__modified_fit_1.32_windows_intelx86__opencl_nvidia.exe
milkyway_separation__modified_fit_1.32_windows_intelx86.exe

Got the same :

ATI based Host
Milkyway@Home 9/12/2014 7:07:33 PM Started download of milkyway_separation__modified_fit_1.32_windows_x86_64__opencl_amd_ati.exe
Milkyway@Home 9/12/2014 7:07:34 PM Giving up on download of milkyway_separation__modified_fit_1.32_windows_x86_64__opencl_amd_ati.exe: permanent HTTP error

NVidia based host
Milkyway@Home 9/12/2014 6:51:02 PM Started download of milkyway_separation__modified_fit_1.32_windows_x86_64__opencl_nvidia.exe
Milkyway@Home 9/12/2014 6:51:03 PM Giving up on download of milkyway_separation__modified_fit_1.32_windows_x86_64__opencl_nvidia.exe: permanent HTTP error

---

I am getting a lot of errors this week:

827071598 617027589 12 Sep 2014, 17:18:49 UTC 12 Sep 2014, 17:19:44 UTC Error while downloading 0.00 0.00 --- Milkyway@Home Separation (Modified Fit) v1.32 (opencl_nvidia)
826699374 616805629 12 Sep 2014, 2:55:36 UTC 12 Sep 2014, 16:44:48 UTC Error while computing 1.05 0.02 --- Milkyway@Home Separation (Modified Fit) v1.30 (opencl_nvidia)

http://milkyway.cs.rpi.edu/milkyway/show_host_detail.php?hostid=561866
http://milkyway.cs.rpi.edu/milkyway/results.php?hostid=561866&offset=0&show_names=0&state=6&appid=

826903403 616930176 12 Sep 2014, 11:15:52 UTC 12 Sep 2014, 16:24:40 UTC Error while computing 1.03 0.01 --- Milkyway@Home Separation (Modified Fit) v1.30 (opencl_nvidia)

http://milkyway.cs.rpi.edu/milkyway/show_host_detail.php?hostid=581205
http://milkyway.cs.rpi.edu/milkyway/results.php?hostid=581205&offset=0&show_names=0&state=6&appid=

16TestStars never worked "Error while computing"

Okay we are looking into the download errors. We restarted the boinc daemons and it started causing some issues with Modfit downloads.

Jake W