Android Beta testing

Message boards : SETI@home Enhanced : Android Beta testing
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Jim
Volunteer tester

Send message
Joined: 6 Sep 13
Posts: 8
Credit: 532
RAC: 0
United States
Message 46782 - Posted: 6 Sep 2013, 23:57:13 UTC

My first WU on my ASUS android tablet (TF101) loaded and seems to run just fine except:
WU 04mr09ab.11676.16841.4.16.235.vlar_1 ran for 15hr 42 min and reached 2050.8% complete. I decided it was not going to stop and aborted the WU.

The software either has a problem detecting the completion of the WU or can't calculate what it has done or how far it has completed.
ID: 46782 · Report as offensive
Alex Storey
Volunteer tester
Avatar

Send message
Joined: 10 Feb 12
Posts: 107
Credit: 305,151
RAC: 0
Greece
Message 46783 - Posted: 7 Sep 2013, 0:11:08 UTC

I wonder what those minus percentages are on that task...
ID: 46783 · Report as offensive
Jim
Volunteer tester

Send message
Joined: 6 Sep 13
Posts: 8
Credit: 532
RAC: 0
United States
Message 46785 - Posted: 7 Sep 2013, 2:54:29 UTC

Good question. I don't have a clue what or why the minus percentages.
I am watching the second one run. It is counting up much slower, about 1% per hour. So I expect it will take several days to run.

ID: 46785 · Report as offensive
Jim
Volunteer tester

Send message
Joined: 6 Sep 13
Posts: 8
Credit: 532
RAC: 0
United States
Message 46797 - Posted: 7 Sep 2013, 16:49:57 UTC

My second WU finished without any problems.
It appears to be a V7 V7.07 (armv7a) application. At 4 1/2 hrs it was only a couple percent done. Later it showed it had run nearly 9 hours and completed.


The first and aborted WU was an V7 V7.09 (armv6) application and it was the one that showed negative percentages done and went to over 2000% done before I aborted it.

ASUS tablet, TF101.
ID: 46797 · Report as offensive
Jim
Volunteer tester

Send message
Joined: 6 Sep 13
Posts: 8
Credit: 532
RAC: 0
United States
Message 46819 - Posted: 9 Sep 2013, 5:52:34 UTC

My forth WU (04mr09ab.18951.18477.4.16.162.vlar_5 ) ran for over 23 hours and reached over 6380 trillion percent complete. I aborted the thing.
All other computers for this WU also failed.

I only errors I am experiencing with the Android application is that it fails to stop or detect an error if the WU fails to end correctly.

ID: 46819 · Report as offensive
Sp@ceNv@der
Volunteer tester

Send message
Joined: 6 Sep 13
Posts: 1
Credit: 5,482
RAC: 0
Message 46880 - Posted: 11 Sep 2013, 16:04:51 UTC

As long as SETI@home v7 v7.07 (armv7-*****) work was running, computing went well without errors. However I've noticed nothing but computing errors since v7 v7.10 started to appear. I'm not gonna keep on trashing work so I've set Boinc to NNT until further notice.

Below: always the same details when looking at the task details

Stderr output

<core_client_version>7.2.9</core_client_version>
<![CDATA[
<message>
process exited with code 193 (0xc1, -63)
</message>
<stderr_txt>
SIGILL: illegal instruction

Exiting...

</stderr_txt>
]]>
_________
ID: 46880 · Report as offensive
Ruud van der Kroef
Volunteer tester

Send message
Joined: 2 Jan 07
Posts: 12
Credit: 7,165,507
RAC: 0
Netherlands
Message 46897 - Posted: 12 Sep 2013, 15:27:23 UTC

Hi,

I just finished my first Seti@home v7 v7.07 WU on my Galaxy S3.
It took the little thing 136,775 CPU seconds (3+ days) to finish. It gave me almost 100 credits.
Hmm, an Einstein BRPS V1.06 only takes about 37,000 CPU seconds (~25%) and gives me 62.50 credits. I am not really a credit addict, but I think I will go for Einstein. Will wait for the second WU to finish.
ID: 46897 · Report as offensive
Jim
Volunteer tester

Send message
Joined: 6 Sep 13
Posts: 8
Credit: 532
RAC: 0
United States
Message 46907 - Posted: 12 Sep 2013, 21:36:11 UTC
Last modified: 12 Sep 2013, 21:57:45 UTC

I had another WU: 04mr09ab.17290.4162.5.16.240 fail to finish. The percent complete said INFINITY after about 10 1/2 hours.
I aborted this WU. As long as the unit progresses toward 100% or something near that I consider it is working correctly. Once it gets into the hundreds/thousands percent complete then I abort the thing.

I been running the Beta testing for a week now, haven't had anything complete/finish successfully. Got a whole 11 points. Mush be in it for the science part of the deal.
Anyway hope this helps.

Upon review of my tasks I find that I did have one WU completely finish and report in correctly: 04mr09ab.10537.20113.4.16.151_2
It ran for over 20 hours and found the following:
Flopcounter: 20283976199634.351562

Spike count: 3
Autocorr count: 2
Pulse count: 25
Triplet count: 0
Gaussian count: 0
19:36:30 (29758): called boinc_finish

The project marked it as invalid and awarded ZERO credits.

I am running the V7 7.10 application, on a ASUS TF101 Transformer tablet. This WU was under armv6-vfp
ID: 46907 · Report as offensive
Geoffk
Volunteer tester

Send message
Joined: 12 Sep 13
Posts: 1
Credit: 287
RAC: 0
Message 46911 - Posted: 12 Sep 2013, 21:47:45 UTC

Advanced preferences for Android include Daily Transfer Limit. Need some idea of mbs needed to get new task and send back results so you can set realistic limit.

Geoffk
ID: 46911 · Report as offensive
arkayn
Volunteer tester
Avatar

Send message
Joined: 16 Jan 07
Posts: 155
Credit: 194,400
RAC: 0
United States
Message 46912 - Posted: 12 Sep 2013, 23:23:35 UTC - in response to Message 46911.  

Advanced preferences for Android include Daily Transfer Limit. Need some idea of mbs needed to get new task and send back results so you can set realistic limit.

Geoffk


AP work units are 8 MB, the return is 350 kilobytes.
ID: 46912 · Report as offensive
Josef W. Segur
Volunteer tester

Send message
Joined: 14 Oct 05
Posts: 1137
Credit: 1,848,733
RAC: 0
United States
Message 46919 - Posted: 13 Sep 2013, 1:34:16 UTC - in response to Message 46911.  

Advanced preferences for Android include Daily Transfer Limit. Need some idea of mbs needed to get new task and send back results so you can set realistic limit.

Geoffk

SETI v7 WUs are ~365 KB, result files range from ~22 KB to ~50 KB in size, depending on how many signals, and what kind, are found.
                                                                   Joe
ID: 46919 · Report as offensive
Jim
Volunteer tester

Send message
Joined: 6 Sep 13
Posts: 8
Credit: 532
RAC: 0
United States
Message 46938 - Posted: 13 Sep 2013, 21:39:19 UTC

It looks like my ASUS TF-101 android can complete a WU without errors. During the last couple days one ran for about 20,000 seconds failed with an unknown error code of 1. The second on ran for 86,000 seconds, completed and validated good.
04mr09ab.17290.4571.5.16.240_4

It ran under v7 7.10 (armv6-vfp)

<core_client_version>7.2.9</core_client_version>
<![CDATA[
<stderr_txt>
setiathome_v7 7.10 Revision: 1976 arm-linux-androideabi-g++ (GCC) 4.6 20120106 (prerelease)
libboinc: BOINC 7.2.4

I even got 82.01 credits.
ID: 46938 · Report as offensive
Profile Siran d'Vel'nahr
Volunteer tester
Avatar

Send message
Joined: 16 Jun 05
Posts: 56
Credit: 234,234
RAC: 0
United States
Message 46950 - Posted: 14 Sep 2013, 20:30:49 UTC

Greetings,

Since this thread is labeled 'Android Beta Testing', I figured this would be an appropriate thread for this.

We have 2:
Lenovo IdeaTab S6000-F tablets
CPU: MTK 8125 QC 1.2 GHz
RAM: 1G LPDDR2
Android: v4.2

Here's the thing: We both have credit for the first WU finished; verified through our accounts. Our tablets are showing 'Credit 0 (on this device) xxxx (total)' on the project list page. Is this a feature that will be enabled later in beta or is this a glitch? 2 inquiring minds would like to know. :)

Thanks in advance... :)

Keep on BOINCing...! :)

CAPT Siran d'Vel'nahr XO
USS Vre'kasht NCC-33187

Siran's website: [ ONLINE! ]
ID: 46950 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 29 May 06
Posts: 1037
Credit: 8,440,339
RAC: 0
United Kingdom
Message 46951 - Posted: 14 Sep 2013, 20:58:50 UTC - in response to Message 46950.  

Here's the thing: We both have credit for the first WU finished; verified through our accounts. Our tablets are showing 'Credit 0 (on this device) xxxx (total)' on the project list page. Is this a feature that will be enabled later in beta or is this a glitch? 2 inquiring minds would like to know. :)

Your host hasn't contacted the project since the Wu validated:

Last contact: 14 Sep 2013, 9:37:13 UTC

Time reported or deadline: 14 Sep 2013, 9:37:13 UTC

How is Boinc on that host supposed to know that it's had validated work if it doesn't contact the project?

Claggy
ID: 46951 · Report as offensive
Jim
Volunteer tester

Send message
Joined: 6 Sep 13
Posts: 8
Credit: 532
RAC: 0
United States
Message 46959 - Posted: 15 Sep 2013, 6:14:52 UTC

Still having problems with WU's not calculating the percent done and completing the WU:
04mr09ab.22422.9888.5.16.167 showed Infinity% complete around 7 hours. I let it run for a total of 20 hours then aborted the WU:
The next one finished/aborted in less than 2 seconds.
ID: 46959 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 29 May 06
Posts: 1037
Credit: 8,440,339
RAC: 0
United Kingdom
Message 46970 - Posted: 15 Sep 2013, 10:24:56 UTC - in response to Message 46959.  

Still having problems with WU's not calculating the percent done and completing the WU:
04mr09ab.22422.9888.5.16.167 showed Infinity% complete around 7 hours. I let it run for a total of 20 hours then aborted the WU:
The next one finished/aborted in less than 2 seconds.

Expect Wu's to take up to 36 hours, aborting Wu's at 20 hours is too soon, your one good recent Wu took 24 hours.

On a side issue, other completed Wu's have come up inconclusive or invalid with overflows (your host is finding more signals than your wingman),
you might want to keep an eye on how warm it's getting, and try putting it on a metal surface to conduct the heat away,

Claggy
ID: 46970 · Report as offensive
Profile Siran d'Vel'nahr
Volunteer tester
Avatar

Send message
Joined: 16 Jun 05
Posts: 56
Credit: 234,234
RAC: 0
United States
Message 46971 - Posted: 15 Sep 2013, 11:39:39 UTC - in response to Message 46951.  

Here's the thing: We both have credit for the first WU finished; verified through our accounts. Our tablets are showing 'Credit 0 (on this device) xxxx (total)' on the project list page. Is this a feature that will be enabled later in beta or is this a glitch? 2 inquiring minds would like to know. :)

Your host hasn't contacted the project since the Wu validated:

Last contact: 14 Sep 2013, 9:37:13 UTC

Time reported or deadline: 14 Sep 2013, 9:37:13 UTC

How is Boinc on that host supposed to know that it's had validated work if it doesn't contact the project?

Claggy

Well, I guess I will just have to accept that. It looks to me like contact will be made about every 2.X days since one WU's work is glacial compared to the other. I don't know enough about tablets yet to understand why one runs fast and the second not when the ARM is a 4 core CPU. I would surmise that they would both get equal billing.

Which was gong to be my second informational request. Why do the 2 WUs run at such greatly different speeds? I have looked in settings and found nothing related.

Inquiring minds and all that... :)

Keep on BOINCing...! :)

CAPT Siran d'Vel'nahr XO
USS Vre'kasht NCC-33187

Siran's website: [ ONLINE! ]
ID: 46971 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 29 May 06
Posts: 1037
Credit: 8,440,339
RAC: 0
United Kingdom
Message 46972 - Posted: 15 Sep 2013, 12:13:55 UTC - in response to Message 46971.  
Last modified: 15 Sep 2013, 12:16:02 UTC

I don't know enough about tablets yet to understand why one runs fast and the second not when the ARM is a 4 core CPU. I would surmise that they would both get equal billing.

Which was gong to be my second informational request. Why do the 2 WUs run at such greatly different speeds? I have looked in settings and found nothing related.

They both have the same AR, but are using different applications, one is on v7.10 (armv7-vfpv3), the other is on v7.10 (armv7-vfpv3d16),
and your completed Wu was a Shortie.

Claggy
ID: 46972 · Report as offensive
Profile Siran d'Vel'nahr
Volunteer tester
Avatar

Send message
Joined: 16 Jun 05
Posts: 56
Credit: 234,234
RAC: 0
United States
Message 46987 - Posted: 16 Sep 2013, 10:07:58 UTC - in response to Message 46972.  

I don't know enough about tablets yet to understand why one runs fast and the second not when the ARM is a 4 core CPU. I would surmise that they would both get equal billing.

Which was gong to be my second informational request. Why do the 2 WUs run at such greatly different speeds? I have looked in settings and found nothing related.

They both have the same AR, but are using different applications, one is on v7.10 (armv7-vfpv3), the other is on v7.10 (armv7-vfpv3d16),
and your completed Wu was a Shortie.

Claggy

My apologies Claggy, but I beg to differ on this...

Upon looking at my task list I see that my first 2 WUs, that are now completed by the way, were running on the same application: SETI@home v7 v7.10 (armv7-vfpv3d16). The 2 WUs currently on my tablet are indicated as running on 2 different applications. One is: SETI@home v7 v7.10 (armv7-neon) and the other: SETI@home v7 v7.10 (armv7-vfpv3).

As I see it, that does not explain why 2 WUs are running at such greatly different speeds on my tablet if running on the same application. There must be another explanation. Perhaps one core is throttled back to help prevent the CPU from burning out? That sounds plausible to me. What do you think?

Keep on BOINCing...! :)

CAPT Siran d'Vel'nahr XO
USS Vre'kasht NCC-33187

Siran's website: [ ONLINE! ]
ID: 46987 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 29 May 06
Posts: 1037
Credit: 8,440,339
RAC: 0
United Kingdom
Message 46993 - Posted: 16 Sep 2013, 16:31:59 UTC - in response to Message 46987.  
Last modified: 16 Sep 2013, 17:12:18 UTC

They both have the same AR, but are using different applications, one is on v7.10 (armv7-vfpv3), the other is on v7.10 (armv7-vfpv3d16),
and your completed Wu was a Shortie.

Claggy

My apologies Claggy, but I beg to differ on this...

The two Wu's I was looking at that you had running at that time were one armv7-vfpv3d16 Wu, and one armv7-vfpv3 Wu, which both have the same AR (look at your wingmen's results) At no point did you say which Wu's you were comparing,

And i repeat, The first completed Wu That You Have Is a 'SHORTIE', As in a DIFFERENT ANGLE RANGE, So has a different set of searches carried out on it.

Claggy
ID: 46993 · Report as offensive
1 · 2 · Next

Message boards : SETI@home Enhanced : Android Beta testing


 
©2021 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.