Posts by Claggy

1) Message boards : Wish List : 64-bit GPU application (Message 134290)
Posted 3 days ago by Claggy
Under Win you could simply run the 32 Bit executable in a 64 Bit OS. I don't know Macs.. is this not possible under OS X?

MrS

With Cuda 6.5, Nvidia is deprecating Support for 32bit Cuda on 64bit OS's, as well as Support for sm_10, sm_11, sm_12, and sm_13 Architectures,
along with many other changes:

CUDA 6.5 MAC OSX

CUDA Toolkit v6.5 Unsupported Features

So expect a lot more 64bit Cuda apps soon.

Claggy
2) Message boards : Problems and Bug Reports : Display issue with Scheduler Log (Message 134214)
Posted 8 days ago by Claggy
Sorry, no one of the E@H team is using IE (all on Firefox, Chrome or Safari). Should work now - does it?

BM

Yes, it is correctly shown in IE11 now:

2014-10-16 13:12:18.7665 [PID=3487] Request: [USER#xxxxx] [HOST#11635026] [IP xxx.xxx.xxx.177] client 7.4.22
2014-10-16 13:12:18.7669 [PID=3487 ] [debug] [HOST#11635026] Resetting nresults_today
2014-10-16 13:12:18.7671 [PID=3487 ] [send] effective_ncpus 1 max_jobs_on_host_cpu 999999 max_jobs_on_host 999999
2014-10-16 13:12:18.7671 [PID=3487 ] [send] effective_ngpus 0 max_jobs_on_host_gpu 999999
2014-10-16 13:12:18.7671 [PID=3487 ] [send] Not using matchmaker scheduling; Not using EDF sim
2014-10-16 13:12:18.7671 [PID=3487 ] [send] CPU: req 28.21 sec, 0.00 instances; est delay 0.00
2014-10-16 13:12:18.7671 [PID=3487 ] [send] work_req_seconds: 28.21 secs
2014-10-16 13:12:18.7671 [PID=3487 ] [send] available disk 0.49 GB, work_buf_min 345600
2014-10-16 13:12:18.7671 [PID=3487 ] [send] active_frac 0.999922 on_frac 0.968772 DCF 2.021718
2014-10-16 13:12:18.7678 [PID=3487 ] [send] [HOST#11635026] is reliable
2014-10-16 13:12:18.7680 [PID=3487 ] [send] set_trust: error rate 0.077377 > 0.050000, don't trust
2014-10-16 13:12:18.7680 [PID=3487 ] [mixed] sending non-locality work first (0.0536)
2014-10-16 13:12:18.7967 [PID=3487 ] [version] no app version available: APP#27 (hsgamma_FGRP4) PLATFORM#11 (arm-unknown-linux-gnueabihf) min_version 0
2014-10-16 13:12:18.7969 [PID=3487 ] [version] no app version available: APP#25 (einsteinbinary_BRP4G) PLATFORM#11 (arm-unknown-linux-gnueabihf) min_version 0
2014-10-16 13:12:18.7971 [PID=3487 ] [version] Checking plan class 'NEON'
2014-10-16 13:12:18.7990 [PID=3487 ] [version] reading plan classes from file '/BOINC/projects/EinsteinAtHome/plan_class_spec.xml'
2014-10-16 13:12:18.7990 [PID=3487 ] [version] CPU lacks feature ' neon '
2014-10-16 13:12:18.7990 [PID=3487 ] [version] Best version of app einsteinbinary_BRP4 is ID 472 (0.21 GFLOPS)
2014-10-16 13:12:18.7990 [PID=3487 ] [send] [HOST#11635026] [WU#201275526 p2030.20131125.G177.70+00.54.C.b3s0g0.00000_1056] using delay bound 1209600 (opt: 1209600 pess: 1209600)
2014-10-16 13:12:18.8004 [PID=3487 ] [debug] Sorted list of URLs follows [host timezone: UTC+0]
2014-10-16 13:12:18.8004 [PID=3487 ] [debug] zone=+03600 url=http://einstein2.aei.uni-hannover.de
2014-10-16 13:12:18.8004 [PID=3487 ] [debug] zone=-18900 url=http://einstein-dl.syr.edu
2014-10-16 13:12:18.8004 [PID=3487 ] [debug] zone=-21600 url=http://einstein-dl2.phys.uwm.edu
2014-10-16 13:12:18.8004 [PID=3487 ] [debug] zone=-28800 url=http://einstein.ligo.caltech.edu
2014-10-16 13:12:18.8006 [PID=3487 ] [send] [HOST#11635026] Sending app_version 472 einsteinbinary_BRP4 11 106 ; 0.21 GFLOPS
2014-10-16 13:12:18.8027 [PID=3487 ] [send] est. duration for WU 201275526: unscaled 82638.89 scaled 172471.50
2014-10-16 13:12:18.8027 [PID=3487 ] [HOST#11635026] Sending [RESULT#460834904 p2030.20131125.G177.70+00.54.C.b3s0g0.00000_1056_1] (est. dur. 172471.50 seconds, delay 1209600, deadline 1414674738)
2014-10-16 13:12:18.8043 [PID=3487 ] [send] don't need more work
2014-10-16 13:12:18.8043 [PID=3487 ] [mixed] sending locality work second
2014-10-16 13:12:18.8060 [PID=3487 ] Sending reply to [HOST#11635026]: 1 results, delay req 60.00
2014-10-16 13:12:18.8071 [PID=3487 ] Scheduler ran 0.044 seconds


Claggy
3) Message boards : News : Einstein@Home scheduler URL changed (Message 134204)
Posted 9 days ago by Claggy
Not Alpha/Beta ...

I beg to differ, Boinc 5.10.28 was announced on the Boinc alpha list for testing on 29th Oct 2007, It was not a recommended release,
Boinc 5.10.45 was the recommended release for the 5.10.x line, anything before that will have been Alpha builds for testing the latest fixes, before moving onto the next build:

•[boinc_alpha] BOINC 5.10.28 released for Windows and Windows x64


Howdy Folks,



This release contains a bug fix for systems that have
problems uninstalling a previous version of BOINC. The new installer
should ignore errors from the older uninstaller and attempt to install
the new BOINC anyway. Another change involves being able to specify how
large your log files can get.



As always log bugs to http://boinc.berkeley.edu/trac and
log test results to http://boinc.berkeley.edu/alpha.




----- Rom



Change Log:

- client: add <max_stdout_file_size> and <max_stderr_file_size>

cc_config.xml options to change log file size limit

from the default of 2 MB

Mac: Fix bugs in scripts for building wxMac-2.6.3; fix XCode project


when using debug build of wxMac-2.6.3.

- MGR: Fixed broken URLs in the Help menu. The URLs are only broken
if the

Manager was built with the wxWidgets unicode variant.


Boinc 5.10.x was finally released on 11th March 2008:

•[boinc_alpha] BOINC 5.10.45 released as the new stable build for all platforms


BOINC 5.10.45 is now the new stable release for all platforms.



This release is a bug fix release which includes fixes for better proxy
support through the use of the <force_auth> configuration option in
cc_config.xml which is defined here:

http://boinc.berkeley.edu/trac/wiki/ClientMessages



This release also contains fixes for startup/shutdown issues when
running on Windows Vista.



Enjoy.



----- Rom


Claggy
4) Message boards : News : Einstein@Home scheduler URL changed (Message 134202)
Posted 9 days ago by Claggy
You seem to foreward from the old scheduler URL to the new one but for some reason this results in this message :

"error in request message: no start tag"

so the foreward could as well be dropped.

Boinc 5.10.28, that dates from 2007, probably an Alpha/Beta version, there's been at least one Curl Bug found since then, Boinc 5.10.30 had two Curl fixes, (there's been others since):

- DEP: Update libCurl to 7.17.1

- Fix build break introduced by new libcurl headers on Windows.

BOINC 5 versions Change Log

It is really time you upgraded, Boinc 7.2.42 is the latest Recommended and has very few problems.

Claggy
5) Message boards : Problems and Bug Reports : Display issue with Scheduler Log (Message 134195)
Posted 9 days ago by Claggy
It's the same on Win 7 and IE11:

2014-10-16 07:33:46.7702 [PID=2044] Request: [USER#xxxxx] [HOST#11667747] [IP xxx.xxx.xxx.177] client 7.2.47 2014-10-16 07:33:46.7708 [PID=2044 ] [send] effective_ncpus 2 max_jobs_on_host_cpu 999999 max_jobs_on_host 999999 2014-10-16 07:33:46.7708 [PID=2044 ] [send] effective_ngpus 0 max_jobs_on_host_gpu 999999 2014-10-16 07:33:46.7708 [PID=2044 ] [send] Not using matchmaker scheduling; Not using EDF sim 2014-10-16 07:33:46.7708 [PID=2044 ] [send] CPU: req 49.05 sec, 0.00 instances; est delay 0.00 2014-10-16 07:33:46.7708 [PID=2044 ] [send] work_req_seconds: 49.05 secs 2014-10-16 07:33:46.7709 [PID=2044 ] [send] available disk 6.64 GB, work_buf_min 172800 2014-10-16 07:33:46.7709 [PID=2044 ] [send] active_frac 0.999947 on_frac 0.849391 DCF 2.856389 2014-10-16 07:33:46.7717 [PID=2044 ] [send] [HOST#11667747] is reliable 2014-10-16 07:33:46.7718 [PID=2044 ] [send] set_trust: error rate 0.054035 > 0.050000, don't trust 2014-10-16 07:33:46.7718 [PID=2044 ] [mixed] sending non-locality work first (0.0179) 2014-10-16 07:33:46.8007 [PID=2044 ] [version] no app version available: APP#27 (hsgamma_FGRP4) PLATFORM#11 (arm-unknown-linux-gnueabihf) min_version 0 2014-10-16 07:33:46.8008 [PID=2044 ] [version] no app version available: APP#25 (einsteinbinary_BRP4G) PLATFORM#11 (arm-unknown-linux-gnueabihf) min_version 0 2014-10-16 07:33:46.8009 [PID=2044 ] [version] Checking plan class 'NEON' 2014-10-16 07:33:46.8027 [PID=2044 ] [version] reading plan classes from file '/BOINC/projects/EinsteinAtHome/plan_class_spec.xml' 2014-10-16 07:33:46.8027 [PID=2044 ] [version] plan class ok 2014-10-16 07:33:46.8027 [PID=2044 ] [version] Best version of app einsteinbinary_BRP4 is ID 502 (0.51 GFLOPS) 2014-10-16 07:33:46.8027 [PID=2044 ] [send] [HOST#11667747] [WU#201275178 p2030.20131125.G177.70+00.54.C.b3s0g0.00000_708] using delay bound 1209600 (opt: 1209600 pess: 1209600) 2014-10-16 07:33:46.8041 [PID=2044 ] [debug] Sorted list of URLs follows [host timezone: UTC+0] 2014-10-16 07:33:46.8041 [PID=2044 ] [debug] zone=+03600 url=http://einstein2.aei.uni-hannover.de 2014-10-16 07:33:46.8041 [PID=2044 ] [debug] zone=-18900 url=http://einstein-dl.syr.edu 2014-10-16 07:33:46.8041 [PID=2044 ] [debug] zone=-21600 url=http://einstein-dl2.phys.uwm.edu 2014-10-16 07:33:46.8041 [PID=2044 ] [debug] zone=-28800 url=http://einstein.ligo.caltech.edu 2014-10-16 07:33:46.8043 [PID=2044 ] [send] [HOST#11667747] Sending app_version 502 einsteinbinary_BRP4 11 142 NEON; 0.51 GFLOPS 2014-10-16 07:33:46.8074 [PID=2044 ] [send] est. duration for WU 201275178: unscaled 34333.62 scaled 115465.52 2014-10-16 07:33:46.8074 [PID=2044 ] [HOST#11667747] Sending [RESULT#460834207 p2030.20131125.G177.70+00.54.C.b3s0g0.00000_708_0] (est. dur. 115465.52 seconds, delay 1209600, deadline 1414654426) 2014-10-16 07:33:46.8097 [PID=2044 ] [send] don't need more work 2014-10-16 07:33:46.8097 [PID=2044 ] [mixed] sending locality work second 2014-10-16 07:33:46.8118 [PID=2044 ] Sending reply to [HOST#11667747]: 1 results, delay req 60.00 2014-10-16 07:33:46.8129 [PID=2044 ] Scheduler ran 0.047 seconds


Claggy
6) Message boards : Problems and Bug Reports : GPU suspended (Message 133711)
Posted 35 days ago by Claggy
Make sure you haven't set the Activity menu to Suspend GPU, and that it is either set to Use GPU based on preferences, or Use GPU always.

Claggy
7) Message boards : Cruncher's Corner : intel gpu help (Message 133709)
Posted 35 days ago by Claggy
As a point of reference, the Intel GPU's absolutely must have a dummy plug or monitor plugged into one of the onboard video ports on the I/O panel of the motherboard, otherwise they will not be recognized and functional in BOINC.

Not true, you can extend the desktop onto it, and not have a dummy plug, Boinc should still detect and use the Intel GP.

That is right click on the desktop, choose 'screen resolution', go 'detect', chose the display that is on your Intel GPU, you'll have to force it detect the GPU by 'try and connect anyway on: VGA', then on 'multiple displays', chose 'extend these displays', Restart Boinc.

Claggy
8) Message boards : Technical News : Fermi LAT Gamma-ray pulsar search #3 "FGRP3" (Message 133550)
Posted 43 days ago by Claggy
Hi,

I have a question or two specifically.

I am running the following or was running two task for computer 11629291:

What I am asking is, why are both task not listed in the current task late or listed as currently finished and reported. I just wanted to know if this last running task is not needed or currently needs to be trashed for computer #11629291.

You received the tasks, didn't complete them before deadline, the tasks have been sent to someone else's host, they've completed the task, your wingmens results have then been validated, and the task, and results have been purged from the Boinc database.

Claggy
9) Message boards : Problems and Bug Reports : Error with Gravitational Wave S6 Directed Search (CasA) (Message 133377)
Posted 52 days ago by Claggy
Hello. There was a problem with task h1_0991.65_S6Directed__S6CasAf40a_992.8Hz_186_1

After 90, 909 % completing, it was going to close cycle (as I understand) because in few hours it was that 90, 909 % and no more. So I aborted it.

A number of Einstein apps display infrequent progress, eventually it would have reached 100%, and may even have spent some time at 100%, if you'd had more patience.

Claggy
10) Message boards : Cruncher's Corner : BOINC weirdness again (Message 133210)
Posted 59 days ago by Claggy
Seems to me, if a task has to go HP the moment it's downloaded, its deadline is too short, or something is still wonky with priority management.

Or this project doesn't have enough resource share allocated to it for the taks to be completed in time, without going into high priority.

Claggy


Next 10

Home · Your account · Message boards

This material is based upon work supported by the National Science Foundation (NSF) under Grants PHY-1104902, PHY-1104617 and PHY-1105572 and by the Max Planck Gesellschaft (MPG). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the investigators and do not necessarily reflect the views of the NSF or the MPG.

Copyright © 2014 Bruce Allen