𝕏

Posts by Conan

21) Message boards : News : New WUs Generation will be paused (Message 775)
Posted 22 Jan 2016 by Profile Conan
Post:
I noticed you released a small batch of tasks for "Beta Testing Version of D.E.N.I.S Application"

Does this mean:

1. We'll see some more beta testing? (as all the tasks were gobbled up)
Yes, we have launch a small group of WUs, which will be bigger if all of those WUs ends correctly. The beta version will be in continuous develop so we will generate new WUs to test in different users and computer configurations
2. The production app will be called "D.E.N.I.S Application"?
The name isn't definitive but for now yes. DENIS will be the common application with different models inside, this is why it's name isn't the one of the model (as the previous one) In the future, each version will have a post in the web page explaining what it have inside. The beta version([url]http://denis.usj.es/the-project/how-it-works/) of the app is the "Kalium"(http://denis.usj.es/denis-kalium/) version of the app.
3. The previous app is completely retired, never to be seen again?
This version is an improvement of the previous one. The optimized version of the app can be modified to work with this new version of the app because the main optimization was in the model and this part have been unmodified.

Thanks!

S.


I hope that answer you ;) , Best Regards, Joel.


Going on these posts, I take it that I have to remove my app_info.xml file or does it just have to be modified so I can receive the current Beta work? If all it needs is to be modified then what are the correct file names and application name?

Thanks
Conan
22) Message boards : Number crunching : Optimized app ? (Message 677)
Posted 12 Nov 2015 by Profile Conan
Post:
On one machine, from time to time (17 WUs in 5 days), I'm getting this :

<core_client_version>7.6.9</core_client_version>
<![CDATA[
<message>
finish file present too long
</message>
<stderr_txt>
Optimized by Sesef v1.5.5 (SSE3) 2015
OpT:1800000.000000
DT:0.002000
F:50
InT:1799000.000000
09:38:50 (3348): called boinc_finish(0)

</stderr_txt>


And the task fails with the following error : 194 (0xc2) EXIT_ABORTED_BY_CLIENT

On another machine, I also got a couple of these :

<core_client_version>7.6.9</core_client_version>
<![CDATA[
<message>
Fonction incorrecte.
(0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
Optimized by Sesef v1.6.1 (AVX2) 2015
OpT:1800000.000000
DT:0.002000
F:50
InT:1799000.000000

</stderr_txt>


And the task fails with the following error : 1 (0x1) Unknown error number

Does anyone know what's wrong ?

No one ?

I'm still getting some "finish file present too long" errors sometimes ...


G'Day toTOW,

I don't really know what might be at issue but I did notice that the computer you are having the most issues with (it has had 94 errors) is ID: 48314, an i7 with 8 CPUs and 4GB RAM.
Could the amount of RAM be the issue as it has the least memory (your i3 has 12GB and other i7's have up to 16Gb), if the computer runs out of memory and sits there waiting till it can get some it could be generating the long finish files?

Just a thought.

Conan
23) Message boards : Number crunching : Long Running Task (Message 470)
Posted 14 Sep 2015 by Profile Conan
Post:
It is GD_jcarro_20150903001429000000_CRLP_Exp_Design_1_conf_4619_2

See Task 14737515

It is at 9 hours 35 minutes and counting at the moment.

Thanks
Conan


It eventually reported fine after 11 Hours, but gave very poor credit (10 cr/h) probably due to the other returned result only recording 1,655 seconds run time.
Mine took a Run Time 40,465.05 CPU Time 39,788.40 Credit 115.46

No idea why it has taken so long to run, glad it was a successful result but for the effort at least double points would of been nice.

Conan
24) Message boards : Number crunching : Long Running Task (Message 468)
Posted 14 Sep 2015 by Profile Conan
Post:
It is GD_jcarro_20150903001429000000_CRLP_Exp_Design_1_conf_4619_2

See Task 14737515

It is at 9 hours 35 minutes and counting at the moment.

Thanks
Conan
25) Message boards : Number crunching : Long Running Task (Message 465)
Posted 14 Sep 2015 by Profile Conan
Post:
I have a Dennis work unit that has been running for over 9 Hours so far, still using a full core.

Are there a few really long running work units in the current batch?

The longest I have had recently has been 9,500 seconds (2.64 hours).

The work unit is running on a Windows XP 32 bit AMD Phenom 955 computer.

Thanks
Conan
26) Message boards : News : New v1.05 DENIS Application (Message 313)
Posted 17 Jun 2015 by Profile Conan
Post:
(Sorry for my english)

Hello everyone! I choose "DENIS@home" and "Update" in actions of my BOINC manager. Is it mean that my manager get new version, or i do something wrong? :)

Great thanks to developers and scientists of that project!


All your current tasks are version 1.05, so all OK.

Conan
27) Message boards : News : Temporaly stop of new WUs (Message 299)
Posted 17 Jun 2015 by Profile Conan
Post:
Hi!

The problem was solved, now the validator works and the flow of WUs is also working ok.

Best regards, Joel.


Thanks for that, and thanks for the update.

Conan
28) Message boards : News : New v1.04 DENIS Application (Message 267)
Posted 4 Jun 2015 by Profile Conan
Post:
Now you have the new application out will there be a correction for all the 64 Bit work units that passed through the 32 Bit computers and gave such low run times when they should not have run at all?

Were these work units faulty? Or did you get valid results from them?

If they were faulty are they going to be reissued to be re-run and re-validated?

And what happens to all credit that has been granted (there is a lot of it, for example one 2 core host has over 1,900,000 points) for these hosts that ran the wrong software version and gave what appeared to be incorrect results?

I suppose we will have to wait to see if the new software version has actually fixed the issue.

Conan
29) Message boards : Number crunching : Credit New ? (Message 262)
Posted 3 Jun 2015 by Profile Conan
Post:
It looks like there is another very fast 2 core host Host 2617
Between 6 and 40 seconds per work unit.

Conan

A couple more:

http://denis.usj.es/denisathome/show_host_detail.php?hostid=2408

http://denis.usj.es/denisathome/show_host_detail.php?hostid=3113


All so far appear to be 2 core Intel CPU's (Pentium (R) and (D) and a Core 2 Duo), plus they are 32 Bit.
Unknown if that is anything to go by.
They don't get much when paired together, but they get through thousands of work units.

Conan


Yep, we are trying to minimize the number of operations that can do those hosts. At the moment, the faulty host that we have found have a limit of 100 operations per day.

The main problem is that they are getting v1.03 version.

Regards, Joel.


Yes, interesting, as they are not 64 Bit but 32 Bit so they should be failing the work units not processing them.

Conan
30) Message boards : Number crunching : Credit New ? (Message 260)
Posted 2 Jun 2015 by Profile Conan
Post:
It looks like there is another very fast 2 core host Host 2617
Between 6 and 40 seconds per work unit.

Conan

A couple more:

http://denis.usj.es/denisathome/show_host_detail.php?hostid=2408

http://denis.usj.es/denisathome/show_host_detail.php?hostid=3113


All so far appear to be 2 core Intel CPU's (Pentium (R) and (D) and a Core 2 Duo), plus they are 32 Bit.
Unknown if that is anything to go by.
They don't get much when paired together, but they get through thousands of work units.

Conan
31) Message boards : Number crunching : Credit New ? (Message 254)
Posted 30 May 2015 by Profile Conan
Post:
how this computer gets credits ?

http://denis.usj.es/denisathome/results.php?hostid=1867&offset=0&show_names=0&state=4&appid=

run time 14 sec, 20 credits ? o_o



We are trying to isolate those result and see what happen. It may be a bug in x86 windows version of the app. We will also check the validator.

Thanks for the info, we will try to solve it as fast as we can.

Regards, Joel.

Looks like you've got another one ... and it's not fixed in V1.03
http://denis.usj.es/denisathome/results.php?hostid=1389&offset=0&show_names=0&state=4&appid=


I am just curious as to how the WUs get validated, runs for 2 seconds but gets full points, is the returned result correct? Or is it really a validator problem? This one host has over 40,000 valid results, 8,000 pending, a RAC of over 60,000 and only a few errors.
That one host makes my 4 look very poorly, even if I am not running just DENIS but a mix of projects.

Conan


Well I hope you fix something up, as this 2 core Pentium D processor has now increased from 765,000 points to over 1,200,000 in one day.

With the RAC going up from 60,000 to over 96,000 and processed results have moved from 48,000 to 79,000.

A very impressive processor taking less than 3 seconds per work unit.

It is also bringing down my credit due to the low claims this Host makes, so when I am paired with it I get lower credit.

Conan


It looks like there is another very fast 2 core host Host 2617
Between 6 and 40 seconds per work unit.

Conan
32) Message boards : Number crunching : Credit New ? (Message 253)
Posted 30 May 2015 by Profile Conan
Post:
how this computer gets credits ?

http://denis.usj.es/denisathome/results.php?hostid=1867&offset=0&show_names=0&state=4&appid=

run time 14 sec, 20 credits ? o_o



We are trying to isolate those result and see what happen. It may be a bug in x86 windows version of the app. We will also check the validator.

Thanks for the info, we will try to solve it as fast as we can.

Regards, Joel.

Looks like you've got another one ... and it's not fixed in V1.03
http://denis.usj.es/denisathome/results.php?hostid=1389&offset=0&show_names=0&state=4&appid=


I am just curious as to how the WUs get validated, runs for 2 seconds but gets full points, is the returned result correct? Or is it really a validator problem? This one host has over 40,000 valid results, 8,000 pending, a RAC of over 60,000 and only a few errors.
That one host makes my 4 look very poorly, even if I am not running just DENIS but a mix of projects.

Conan


Well I hope you fix something up, as this 2 core Pentium D processor has now increased from 765,000 points to over 1,200,000 in one day.

With the RAC going up from 60,000 to over 96,000 and processed results have moved from 48,000 to 79,000.

A very impressive processor taking less than 3 seconds per work unit.

It is also bringing down my credit due to the low claims this Host makes, so when I am paired with it I get lower credit.

Conan
33) Message boards : Number crunching : Credit New ? (Message 251)
Posted 29 May 2015 by Profile Conan
Post:
how this computer gets credits ?

http://denis.usj.es/denisathome/results.php?hostid=1867&offset=0&show_names=0&state=4&appid=

run time 14 sec, 20 credits ? o_o



We are trying to isolate those result and see what happen. It may be a bug in x86 windows version of the app. We will also check the validator.

Thanks for the info, we will try to solve it as fast as we can.

Regards, Joel.

Looks like you've got another one ... and it's not fixed in V1.03
http://denis.usj.es/denisathome/results.php?hostid=1389&offset=0&show_names=0&state=4&appid=


I am just curious as to how the WUs get validated, runs for 2 seconds but gets full points, is the returned result correct? Or is it really a validator problem? This one host has over 40,000 valid results, 8,000 pending, a RAC of over 60,000 and only a few errors.
That one host makes my 4 look very poorly, even if I am not running just DENIS but a mix of projects.

Conan
34) Message boards : Number crunching : Maximum Elapsed Time Exceeded (Message 228)
Posted 26 May 2015 by Profile Conan
Post:
All my Windows XP 32 bit work units are failing with this error message.

<message>
Maximum elapsed time exceeded
</message>
<stderr_txt>
Operation time:1000000.000000
Diferential Time:0.002000
Frequency:50
Initial Time:999000.000000


Unhandled Exception Detected...

- Unhandled Exception Record -
Reason: Breakpoint Encountered (0x80000003) at address 0x7c90120e

See WU 4064725
Also WU 4071990

So far 64 bit Linux seems unaffected.

Conan
35) Message boards : News : DataBase Maintenance (Message 173)
Posted 13 May 2015 by Profile Conan
Post:
No worries, thanks for the heads up.

Conan


Previous 20