1) Message boards : Sixtrack Application : SIXTRACKTEST (Message 34340)
Posted 9 Feb 2018 by BetelgeuseFive
Post:
I just noted one of my systems received a number of tasks the day before yesterday.
Unfortunately all of them failed:

</stderr_txt>
<message>
upload failure: <file_xfer_error>
<file_name>wTestNew_hl13B1_20180121__1__s__62.31_60.32__2_4__4__85.2631_1_sixvf_boinc334_1_r628128846_2</file_name>
<error_code>-131 (file size too big)</error_code>
</file_xfer_error>

</message>
]]>

Looks like it is not just me having these problems:

https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=85955467

Tom
2) Message boards : Sixtrack Application : SIXTRACKTEST (Message 34115)
Posted 27 Jan 2018 by BetelgeuseFive
Post:
There seems to be a problem with some of the new workunits as well.
On this host https://lhcathome.cern.ch/lhcathome/results.php?hostid=10497508 I had two tasks that were sent out in the first batch and these workunits are not marked as "WU cancelled" (probably the ones in the message log below that are marked as expired), but they still show as green in the task list on the website.
It looks like two new tasks were issued. One of these is running fine, but the second one has a missing file error during download.

27-Jan-2018 11:26:34 [LHC@home] Requesting new tasks for CPU
27-Jan-2018 11:26:40 [LHC@home] Scheduler request completed: got 1 new tasks
27-Jan-2018 11:26:40 [LHC@home] Didn't resend lost task wTestNew_hl13B1_20180121__1__s__62.31_60.32__14.4_14.5__6__15_1_sixvf_boinc251_1 (expired)
27-Jan-2018 11:26:40 [LHC@home] Didn't resend lost task wTestNew_hl13B1_20180121__1__s__62.31_60.32__11.7_11.8__6__45_1_sixvf_boinc118_0 (expired)
27-Jan-2018 11:26:40 [LHC@home] Resent lost task wTestNew_hl13B1_20180121__1__s__62.31_60.32__2_4__6__9.47368_1_sixvf_boinc282_2
27-Jan-2018 11:26:40 [LHC@home] State file error: missing file wTestNew_hl13B1_20180121__1__s__62.31_60.32__2_4__6__9.47368_1_sixvf_boinc282_2_r1617352886_2
27-Jan-2018 11:26:40 [LHC@home] Can't handle task wTestNew_hl13B1_20180121__1__s__62.31_60.32__2_4__6__9.47368_1_sixvf_boinc282_2 in scheduler reply

Why are the expired tasks still listed in green on the website (and why do I keep receiving the "Didn't resend lost task" messages ?
What is wrong with the second new task ?

Thanks,

Tom
3) Message boards : Sixtrack Application : SIXTRACKTEST (Message 34103)
Posted 26 Jan 2018 by BetelgeuseFive
Post:
Received one, but the website shows four tasks for this host: https://lhcathome.cern.ch/lhcathome/results.php?hostid=10497508
It downloaded without problems and is running right now (it may take a couple of days on my Raspberry Pi model 2B).

Thanks for fixing this,

Tom
4) Message boards : Sixtrack Application : SIXTRACKTEST (Message 34049)
Posted 23 Jan 2018 by BetelgeuseFive
Post:
There seems to be a problem downloading the the new sixtracktest tasks (same problem on different systems):

23-Jan-2018 12:29:10 [LHC@home] Sending scheduler request: To fetch work.
23-Jan-2018 12:29:10 [LHC@home] Requesting new tasks for CPU
23-Jan-2018 12:29:13 [LHC@home] Scheduler request completed: got 2 new tasks
23-Jan-2018 12:29:13 [LHC@home] Resent lost task wTestNew_hl13B1_20180121__1__s__62.31_60.32__14.4_14.5__6__15_1_sixvf_boinc251_1
23-Jan-2018 12:29:13 [LHC@home] Resent lost task wTestNew_hl13B1_20180121__1__s__62.31_60.32__11.7_11.8__6__45_1_sixvf_boinc118_0
23-Jan-2018 12:29:13 [LHC@home] State file error: missing file wTestNew_hl13B1_20180121__1__s__62.31_60.32__14.4_14.5__6__15_1_sixvf_boinc251_1_r110204253_2
23-Jan-2018 12:29:13 [LHC@home] Can't handle task wTestNew_hl13B1_20180121__1__s__62.31_60.32__14.4_14.5__6__15_1_sixvf_boinc251_1 in scheduler reply
23-Jan-2018 12:29:13 [LHC@home] State file error: missing file wTestNew_hl13B1_20180121__1__s__62.31_60.32__11.7_11.8__6__45_1_sixvf_boinc118_0_r3465899_2
23-Jan-2018 12:29:13 [LHC@home] Can't handle task wTestNew_hl13B1_20180121__1__s__62.31_60.32__11.7_11.8__6__45_1_sixvf_boinc118_0 in scheduler reply

Tom
5) Message boards : Sixtrack Application : AVX Sixtrack version (Message 33955)
Posted 20 Jan 2018 by BetelgeuseFive
Post:
On my i5-6500 system I received several AVX tasks a couple of days ago.
The application details show:

SixTrack 46.30 windows_x86_64 (sse2)

Number of tasks completed 958
Consecutive valid tasks 411
Average processing rate 15.29 GFLOPS

SixTrack 46.30 windows_x86_64 (avx):

Number of tasks completed 1
Consecutive valid tasks 3
Average processing rate 59.42 GFLOPS

As the average processing rate for the AVX tasks is higher I would expect to be receiving more AVX work, but since the initial couple of tasks I am receiving only SSE2 tasks again. Does this make sense ?

Tom
6) Message boards : Sixtrack Application : SIXTRACKTEST (Message 33893)
Posted 17 Jan 2018 by BetelgeuseFive
Post:
Could someone give a status update please ?
I would like to have an application for Raspberry Pi (ARM on Linux), but it has not yet been released for sixtrack.
The sixtracktest application seemed to work fine.
Any chance of releasing the sixtracktest ARM Linux binaries for sixtrack (there seems to be plenty of work available there at the moment) ?

Thanks,

Tom
7) Message boards : Sixtrack Application : Transfer issues (Message 33517)
Posted 26 Dec 2017 by BetelgeuseFive
Post:
Unable to upload at the moment; I get an error message I have not seen before:

26/12/2017 15:00:35 | LHC@home | [error] Error reported by file upload server: [w5_hllhc10_sqz1500_Qcol_chr20_w5__6__s__62.31_60.32__18_20__5__22.5_1_sixvf_boinc806_1_r2118365745_0] locked by file_upload_handler PID=-1

Tom
8) Message boards : Sixtrack Application : SIXTRACKTEST (Message 32365)
Posted 9 Sep 2017 by BetelgeuseFive
Post:
Added two Raspberry Pi model 2B's and one Raspberry Pi model 3B.
Tasks take appr. 8 hours on the model 2B and appr. 4 hours on the model 3B.
There is a problem with memory usage however. Running one tasks at the time is fine, but when running multiple tasks pretty soon one of them will be in 'Waiting for memory' state and for some reason Boinc does not start a task from another project.
Here is a list of the 4 Boinc tasks running on one of my Pi's.
There are two BHspin tasks (Universe@home), one SETI@home task and one sixtracktest task. The universe@home tasks take 2.3 Mb, the seti@home task 36.6 Mb, while the sixtracktest task takes 330.2 Mb (RESident set size column).

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND

907 boinc 39 19 457016 330288 1988 R 99.5 34.9 1:09.98 sixtracktest_ar
910 boinc 39 19 38216 36692 2852 R 99.5 3.9 1:05.03 setiathome_8.06
912 boinc 39 19 3780 2372 2120 R 99.5 0.3 0:57.29 BHspin2_1_arm-u
914 boinc 39 19 3780 2312 2060 R 98.2 0.2 0:54.73 BHspin2_1_arm-u

One one of my Pi's (it had another sixtracktest task 'Waiting for memory') I was unable to start BOINC manager (probably because there was not enough memory).

I don't know if this is feasible, but an application using less memory would be useful here ...

Tom
9) Message boards : Sixtrack Application : SIXTRACKTEST (Message 32334)
Posted 7 Sep 2017 by BetelgeuseFive
Post:
Just added a couple of Raspberry Pi's. Hope there will be more work soon ...

Tom
10) Message boards : Number crunching : Host messing up tons of results (Message 27244)
Posted 30 Mar 2015 by BetelgeuseFive
Post:
I am having an inconclusive against this host as well.
Looks like this host finishes each and every task in 2 seconds.
Weird thing is that some of them get validated, even against other tasks that do use (significant) amounts of CPU-time. Examples:

http://lhcathomeclassic.cern.ch/sixtrack/workunit.php?wuid=30270731
http://lhcathomeclassic.cern.ch/sixtrack/workunit.php?wuid=30273145
http://lhcathomeclassic.cern.ch/sixtrack/workunit.php?wuid=30279506
http://lhcathomeclassic.cern.ch/sixtrack/workunit.php?wuid=30275400

Tom

Same problem cause of that host.

I've got inconclusive validation for 2 or 3 ~8h tasks, that means 1 core-day wasted. A bit frustrating.

Thank you for your support.
11) Message boards : Number crunching : Question about inconcusive result (Message 25872)
Posted 3 Oct 2013 by BetelgeuseFive
Post:
Hi folks,

I am new to LHC@home, but have crunched several other projects.
One of my first tasks on LHC returned inconclusive:

http://lhcathomeclassic.cern.ch/sixtrack/workunit.php?wuid=9409507

On other projects I have worked on one of the tasks will be marked as invalid if two other tasks are a match and the workunit is validated. Over here I see that the task remains inconclusive. Is there any reason for this (and is an inconclusive task something I should worry about) ?

Thanks,

Tom



©2024 CERN