Posts by Thomas Krajewsky
log in
1) Message boards : Number crunching : Database error. unable to upload (Message 22162)
Posted 663 days ago by Profile Thomas Krajewsky
I have one box acting normally with connection every 5 mins.
Others with "can't resolve host name", very strange!


The same here:

C:\>nslookup boinc.thesonntags.com 8.8.8.8
Server: google-public-dns-a.google.com
Address: 8.8.8.8

Nicht autorisierende Antwort:
Name: boinc.thesonntags.com


C:\>nslookup boinc.thesonntags.com 145.253.2.11
Server: dns1.arcor-ip.de
Address: 145.253.2.11

Nicht autorisierende Antwort:
Name: boinc.thesonntags.com


C:\>nslookup boinc.thesonntags.com 208.67.220.220
Server: resolver2.opendns.com
Address: 208.67.220.220

Nicht autorisierende Antwort:
Name: boinc.thesonntags.com


C:\>


(But i can reach this forum with firefox on my laptop, where DNS-Cache has an entry for 'boinc.thesonntags.com')

Small workaround: Edit your hosts-file and add
    50.158.27.76 thesonntags.com
    50.158.27.76 boinc.thesonntags.com

until DNS is fixed.

2) Message boards : Windows : 290X second card 'fails' and crashes my system (Message 21630)
Posted 813 days ago by Profile Thomas Krajewsky
Doh! The first tech support question should always be, "Is it plugged in?"

And the second is: "Have you pushed the power button/turned it on?"
Happened in reality when i came to the computer of a colleague at work to fix it.

Short repair time for me. :-)
3) Message boards : Windows : Confirmation errors (Message 21321)
Posted 851 days ago by Profile Thomas Krajewsky
FYI, reduce_cpu=0 is the default in the sieve v1.21 app although you can still set reduce_cpu=1 which will use the CPU to to the summarization but the larger the sieve_size setting, the more work the CPU will then have to do. The items_per_kernel parameter is not supported because sieve_size now controls the items per kernel.

reduce_cpu with '_' is the correct version?
I started without '_', i copied that from your post:
http://boinc.thesonntags.com/collatz/forum_thread.php?id=1273&postid=20542#20542
4) Message boards : Windows : Confirmation errors (Message 21317)
Posted 851 days ago by Profile Thomas Krajewsky
I found the hint about VS2012 C++ runtime update. I have installed SP4 and now the Sieve apps are running.


That's fine, here my .config for sieve wus with GTX 660:
verbose=1
items_per_kernel=20
kernels_per_reduction=48
threads=10
sleep=1
lut_size=15
reduceCPU=0
sieve_size=30

(The values items_per_kernel and reducecpu are currently with V1.21 not supported.)

If your computer gets unusuable, try increasing the sleep value and/or decreasing the other values to your old working ones.

Another solution maybe an option in boinc manager, you can tell it to stop GPU processing for x minutes if you use keyboard or mouse.

Third possible solution: You may use two sets of .config files, one for working, the other for crunching and two batch files for copying the appropiate files into boincs data directory.
5) Message boards : Windows : Confirmation errors (Message 21307)
Posted 852 days ago by Profile Thomas Krajewsky
Since I have changed from 6.04 to 6.06 (CUDA55) I get errors for my computed tasks. The status in my task list is "Bestätigungsfehler" (confirmation error).
Task ids for example: 23016670, 23003597, 22975627.
I am using a NVIDIA GeForce GTX 660 (2048 MB) with Windows 7 SP1.

My cuda55.config is:
verbose=1
threads=8
items_per_kernel=18
kernels_per_reduction=8
sleep=1
lut_size=19

What is going wrong?

And it has become slower. Solo Collatz with 6.04 took about 60 minutes, now it takes 80 minutes.


lut_size = 19 gives 8*2^19=4194304, that's 4MB, the card has only a cache of 384KB

For better values in the .config:
http://boinc.thesonntags.com/collatz/forum_thread.php?id=1009&postid=16503#16503

GeForce GTX 660, Keppler:
960 CUDA cores
384 KB Cache
2048 MB RAM

Try this
verbose=1
items_per_kernel=20
kernels_per_reduction=9
threads=9
sleep=1
lut_size=15


Or try the sieve (beta) apps, they run now very well without validation errors.
6) Message boards : News : Collatz Sieve v1.21 released for Windows (Message 21228)
Posted 865 days ago by Profile Thomas Krajewsky
I assume that the VS2013 or VS2015 C++ runtimes would be backward compatible although I've not tested that. Given that Win 10 works, that is probably the case.


Google, google: Found two MSDN blog entries, so it's somewhat 'Official'.

http://blogs.msdn.com/b/vcblog/archive/2014/06/10/the-great-crt-refactoring.aspx
http://blogs.msdn.com/b/oldnewthing/archive/2014/04/11/10516280.aspx

If i understand the first blog right, they will stop the versioning of the CRT in Visual Studio 2015 (=V14), but for the older releases you will have to distribute the suitable Runtime Libraries.

On Windows 10:
'dir c:\Windows\msvcr*.dll /s' shows me 36 files in different folders, i think these include all the Runtime Libraries MS ever released.
7) Message boards : News : Collatz Sieve v1.21 released for Windows (Message 21220)
Posted 865 days ago by Profile Thomas Krajewsky
So far over 100 valid WUs, no invalid ones, that's fine.

Some information:

Windows 10 doesn't need VC Runtime update, but Windows 7 and Windows 2008 R2 need it.

I had also to reduce the sieve sizes to get it working:
Geforce 820M (2GB RAM) from 30 to 29
GTX 660 and GTX 750 Ti (both 2GB RAM) from 32 to 30

Does the kernel from 1.2x need more RAM than kernel 1.10?
8) Message boards : News : Collatz Sieve 1.20 released for Windows, Linux, and OS X (Message 21199)
Posted 867 days ago by Profile Thomas Krajewsky
invalid
http://boinc.thesonntags.com/collatz/result.php?resultid=21722593
http://boinc.thesonntags.com/collatz/result.php?resultid=21721088

error (no stderr output, state: 998 (0x3e6) Unknown error number)
http://boinc.thesonntags.com/collatz/result.php?resultid=21722905
http://boinc.thesonntags.com/collatz/result.php?resultid=21723123
http://boinc.thesonntags.com/collatz/result.php?resultid=21722162

error
http://boinc.thesonntags.com/collatz/result.php?resultid=21723240
stderr output:
<core_client_version>7.6.9</core_client_version>
<![CDATA[
<message>
Zugriff verweigert (-> german for 'access denied')
(0x5) - exit code 5 (0x5)
</message>
<stderr_txt>
Collatz Conjecture Sieve 1.20 Windows x86_64 for OpenCL
Written by Slicker (Jon Sonntag) of team SETI.USA
Based on the AMD Brook+ kernels by Gipsel of team Planet 3DNow!
Sieve code and OpenCL optimization provided by Sosiris of team BOINC@Taiwan
Unrecognized command: items_per_kernel=20
Unrecognized command: reducecpu=0
Collatz Config Settings:
verbose 1 (yes)
kernels/reduction 48
threads 2^7 (128)
lut_size 13 (65536 bytes)
sieve_size 2^30 (51085096 bytes)
sleep 1
cache_sieve 1 (yes)
reducecpu 0 (no)
Platform NVIDIA
Device 0000007A9194BA30
Max Dimensions 3
Max Work Items 1024 1024 64
Max Work Groups 1024
Max Kernel Threads 1024
Device Vendor NVIDIA Corporation
Name GeForce 820M
Driver Version 355.82
OpenCL Version OpenCL 1.1 CUDA
actual threads 128
Error: (-5)Out of resources at 1641 of RunSteps
Error: GPU steps do not match CPU steps. Workunit processing aborted.
09:30:55 (1452): called boinc_finish

</stderr_txt>
]]>
The .config for this wu was the same as for 1.10, which were running well.
9) Message boards : Web site : Website 'The Knights Who Say 'Ni!'' dead`? (Message 21152)
Posted 875 days ago by Profile Thomas Krajewsky
The Knights Who Say 'Ni!' -> http://stats.kwsn.net/

This statitics website seems to be dead, it shows only a lot of advertisement crap (german language), but nothing related to BOINC projects.

Time to sort it out?


Maybe they are just changing domains, they can get expensive.


Have found this via Google:
http://stats.kwsn.info/

They have closed the stats services, so someone else maybe the new owner of the old domain and is using the old links to generate some traffic for his ads.

http://www.kwsn.info/ is just a usual team site, the stats link there leads to a short list of other stats services.

The registrar information for both domains are different, anomyzed by two services, one in panama, the other in canada.
10) Message boards : Web site : Website 'The Knights Who Say 'Ni!'' dead`? (Message 21136)
Posted 876 days ago by Profile Thomas Krajewsky
The Knights Who Say 'Ni!' -> http://stats.kwsn.net/

This statitics website seems to be dead, it shows only a lot of advertisement crap (german language), but nothing related to BOINC projects.

Time to sort it out?
11) Message boards : News : Collatz Sieve 1.10 Released for Windows (Message 21096)
Posted 882 days ago by Profile Thomas Krajewsky
Visual Studio did not catch the error "if (x = 1)" whereas the old version of the GCC compiler used by OS X knew it should be "if (x == 1)". There's a big difference between assigning a value and comparing a value. That version 11 of a non-free compiler doesn't catch that is a bit unsettling. Looks like new versions may be developed on Linux first from now on.


If your programming language is 'C' it's not an error, it's a feature of the Language! (But not a really good one)
But a warning should come from your compiler, that's true.

Are you doing exercises for the next 'International Obfuscated C Code Contest'? ;-)
http://www.ioccc.org/
12) Message boards : Number crunching : Optimizing Collatz v6.xx OpenCL and CUDA Applications (Message 21080)
Posted 883 days ago by Profile Thomas Krajewsky
Hello Co-Crunchers

Since the last updates of the CUDA-versions my times doubled nearly, so it was time for some optimization work.

So far i got it to run my old times with this configs:

cc_config.xml:
<cc_config>
<log_flags>
<sched_op_debug>1</sched_op_debug>
</log_flags>
<options>
<use_all_gpus>1</use_all_gpus>
</options>
</cc_config>


app_config.xml:
<app_config>
<app>
<name>mini_collatz</name>
<max_concurrent>1</max_concurrent>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>0.5</cpu_usage>
</gpu_versions>
</app>
<app>
<name>solo_collatz</name>
<max_concurrent>1</max_concurrent>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>0.5</cpu_usage>
</gpu_versions>
</app>
<app>
<name>large_collatz</name>
<max_concurrent>1</max_concurrent>
<gpu_versions>
<gpu_usage>1.0</gpu_usage>
<cpu_usage>0.5</cpu_usage>
</gpu_versions>
</app>
</app_config>


http://boinc.thesonntags.com/collatz/show_host_detail.php?hostid=136304
(GeForce GTX 750 Ti, Maxwell: 640 CUDA cores, 2048 KB Cache, 2048 MB RAM)
*.config
verbose=1
items_per_kernel=20
kernels_per_reduction=8
threads=9
sleep=2
lut_size=17


http://boinc.thesonntags.com/collatz/show_host_detail.php?hostid=119238
(GeForce GTX 660, Keppler: 960 CUDA cores, 384 KB Cache, 2048 MB RAM)
*.config
verbose=1
items_per_kernel=20
kernels_per_reduction=9
threads=9
sleep=1
lut_size=15


http://boinc.thesonntags.com/collatz/show_host_detail.php?hostid=164382
(GeForce 820M, Fermi: 96 CUDA cores, 128 KB Cache, 2048 MB RAM)
*.config
verbose=1
items_per_kernel=20
kernels_per_reduction=9
threads=6
sleep=1
lut_size=13


But there are now a greater number of invalid tasks:

Large CUDA
- 1x Valid
- 4x Invalid
= 80% Invalid

Solo CUDA
- 20x Valid
- 5x Invalid
= 20% Invalid

Mini CUDA
- 123x Valid
- 12x Invalid
= ~10% Invalid

What's going wrong, i can't see any reasons in stderr output?
13) Message boards : News : Android 5, aka Lolipop or Android L not getting work (Message 20275)
Posted 1070 days ago by Profile Thomas Krajewsky
My device (http://boinc.thesonntags.com/collatz/show_host_detail.php?hostid=163394) got some work on 12 Feb 2015 but both failed:

http://boinc.thesonntags.com/collatz/result.php?resultid=14411691
http://boinc.thesonntags.com/collatz/result.php?resultid=14411677

Both with the same result:
error: only position independent executables (PIE) are supported.

Today (14 Feb 2015) i got two new ones and it seems they are running well:

http://boinc.thesonntags.com/collatz/result.php?resultid=14464950
http://boinc.thesonntags.com/collatz/result.php?resultid=14458101




Main page · Your account · Message boards


Copyright © 2018 Jon Sonntag; All rights reserved.