Missing Apps
log in

Advanced search

Message boards : News : Missing Apps

1 · 2 · 3 · Next
Author Message
Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 2
Message 19248 - Posted: 27 Apr 2014, 22:59:57 UTC

If there is a specific plan class for a platform that is missing from the micro, mini, or large Collatz applications, please let me know. The solo work units are no longer multi-size so you will need to choose the application within the venue that work best for your machine. If you don't select an application, you should randomly get work from any that have applications for your platform and plan class.

I also increased the feeder to handle 600 work units in shared memory (up from 200), so if you are have problems getting a specific type of work unit, let me know so I can investigate further.

jjwhalen
Avatar
Send message
Joined: 17 Apr 10
Posts: 20
Credit: 202,526,055
RAC: 0
Message 19251 - Posted: 28 Apr 2014, 2:39:31 UTC

Computer ID 23085:

With only Solo Collatz & CUDA nVidia selected, I am unable to get new work, despite Server Status showing 350+ Solo available.

With Solo Mini & Micro selected, I am able to d/l Mini. At least Mini is my second choice.
____________
Best wishes:)

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 2
Message 19253 - Posted: 28 Apr 2014, 2:54:12 UTC - in response to Message 19251.

I removed the multi-size logic from the scheduler as well as changed the server_state status of previously generated solo WUs so you should now be able to get work. Please let me know if you continue to have issues getting solo WUs.

Profile Gandalf_the_Grey
Avatar
Send message
Joined: 2 May 13
Posts: 40
Credit: 336,096,852
RAC: 0
Message 19254 - Posted: 28 Apr 2014, 3:01:32 UTC

FYI -

4/27/2014 7:53:03 PM | | Starting BOINC client version 7.2.42 for windows_x86_64
4/27/2014 7:53:03 PM | | log flags: file_xfer, sched_ops, task, checkpoint_debug, file_xfer_debug
4/27/2014 7:53:03 PM | | log flags: http_xfer_debug, sched_op_debug
4/27/2014 7:53:03 PM | | Libraries: libcurl/7.25.0 OpenSSL/1.0.1 zlib/1.2.6
4/27/2014 7:53:03 PM | | Data directory: C:\ProgramData\BOINC
4/27/2014 7:53:03 PM | | Running under account Gandalf
4/27/2014 7:53:03 PM | | CAL: ATI GPU 0: AMD Radeon HD 7700 series (Capeverde) (CAL version 1.4.1848, 1024MB, 984MB available, 3680 GFLOPS peak)
4/27/2014 7:53:03 PM | | CAL: ATI GPU 1: AMD Radeon HD 7700 series (Capeverde) (CAL version 1.4.1848, 1024MB, 984MB available, 3200 GFLOPS peak)
4/27/2014 7:53:03 PM | | OpenCL: AMD/ATI GPU 0: AMD Radeon HD 7700 series (Capeverde) (driver version 1348.5 (VM), device version OpenCL 1.2 AMD-APP (1348.5), 1024MB, 984MB available, 3680 GFLOPS peak)
4/27/2014 7:53:03 PM | | OpenCL: AMD/ATI GPU 1: AMD Radeon HD 7700 series (Capeverde) (driver version 1348.5 (VM), device version OpenCL 1.2 AMD-APP (1348.5), 1024MB, 984MB available, 3200 GFLOPS peak)
4/27/2014 7:53:03 PM | | OpenCL CPU: Intel(R) Core(TM) i7 CPU 950 @ 3.07GHz (OpenCL driver vendor: Advanced Micro Devices, Inc., driver version 1348.5 (sse2), device version OpenCL 1.2 AMD-APP (1348.5))
4/27/2014 7:53:03 PM | | app version refers to missing GPU type NVIDIA
4/27/2014 7:53:03 PM | Collatz Conjecture | Application uses missing NVIDIA GPU
4/27/2014 7:53:03 PM | | app version refers to missing GPU type NVIDIA
4/27/2014 7:53:03 PM | Collatz Conjecture | Application uses missing NVIDIA GPU

4/27/2014 7:53:03 PM | | Host name: Gandalf
4/27/2014 7:53:03 PM | | Processor: 8 GenuineIntel Intel(R) Core(TM) i7 CPU 950 @ 3.07GHz [Family 6 Model 26 Stepping 5]
4/27/2014 7:53:03 PM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 cx16 sse4_1 sse4_2 popcnt syscall nx lm vmx tm2 pbe
4/27/2014 7:53:03 PM | | OS: Microsoft Windows 7: Ultimate x64 Edition, Service Pack 1, (06.01.7601.00)
4/27/2014 7:53:03 PM | | Memory: 15.99 GB physical, 16.99 GB virtual
4/27/2014 7:53:03 PM | | Disk: 119.24 GB total, 61.08 GB free
4/27/2014 7:53:03 PM | | Local time is UTC -7 hours
4/27/2014 7:53:03 PM | Collatz Conjecture | Found app_config.xml
4/27/2014 7:53:03 PM | Collatz Conjecture | Your app_config.xml file refers to an unknown application 'micro_collatz'. Known applications: 'solo_collatz', 'large_collatz', 'mini_collatz'
4/27/2014 7:53:03 PM | | Config: use all coprocessors
4/27/2014 7:53:03 PM | | Config: GUI RPCs allowed from:
4/27/2014 7:53:03 PM | | 192.168.1.100
4/27/2014 7:53:03 PM | | 192.168.1.101
4/27/2014 7:53:03 PM | | 192.168.1.102
4/27/2014 7:53:03 PM | | 192.168.1.103
4/27/2014 7:53:03 PM | | 192.168.1.104
4/27/2014 7:53:03 PM | Asteroids@home | URL http://asteroidsathome.net/boinc/; Computer ID 22409; resource share 100
4/27/2014 7:53:03 PM | rosetta@home | URL http://boinc.bakerlab.org/rosetta/; Computer ID 1621546; resource share 100
4/27/2014 7:53:03 PM | Poem@Home | URL http://boinc.fzk.de/poem/; Computer ID 160548; resource share 100
4/27/2014 7:53:03 PM | Collatz Conjecture | URL http://boinc.thesonntags.com/collatz/; Computer ID 126742; resource share 100
4/27/2014 7:53:03 PM | Chess@Home | URL http://chess.qugate.org/; Computer ID 307; resource share 100
4/27/2014 7:53:03 PM | Einstein@Home | URL http://einstein.phys.uwm.edu/; Computer ID 7474065; resource share 100
4/27/2014 7:53:03 PM | NFS@Home | URL http://escatter11.fullerton.edu/nfs/; Computer ID 429650; resource share 100
4/27/2014 7:53:03 PM | NumberFields@home | URL http://numberfields.asu.edu/NumberFields/; Computer ID 12300; resource share 100
4/27/2014 7:53:03 PM | OProject@Home | URL http://oproject.info/; Computer ID 23576; resource share 100
4/27/2014 7:53:03 PM | pogs | URL http://pogs.theskynet.org/pogs/; Computer ID 19178; resource share 100
4/27/2014 7:53:03 PM | WUProp@Home | URL http://wuprop.boinc-af.org/; Computer ID 54527; resource share 100
4/27/2014 7:53:03 PM | Enigma@Home | URL http://www.enigmaathome.net/; Computer ID 117480; resource share 100
4/27/2014 7:53:03 PM | yoyo@home | URL http://www.rechenkraft.net/yoyo/; Computer ID 93269; resource share 100
4/27/2014 7:53:03 PM | World Community Grid | URL http://www.worldcommunitygrid.org/; Computer ID 1511352; resource share 100
4/27/2014 7:53:03 PM | Collatz Conjecture | General prefs: from Collatz Conjecture (last modified 27-Apr-2014 01:14:19)
4/27/2014 7:53:03 PM | Collatz Conjecture | Computer location: work
4/27/2014 7:53:03 PM | Collatz Conjecture | General prefs: no separate prefs for work; using your defaults
4/27/2014 7:53:03 PM | | Reading preferences override file
4/27/2014 7:53:03 PM | | Preferences:
4/27/2014 7:53:03 PM | | max memory usage when active: 16375.11MB
4/27/2014 7:53:03 PM | | max memory usage when idle: 16375.11MB
4/27/2014 7:53:03 PM | | max disk usage: 10.00GB
4/27/2014 7:53:03 PM | | (to change preferences, visit a project web site or select Preferences in the Manager)
4/27/2014 7:53:03 PM | | Not using a proxy
4/27/2014 7:53:05 PM | Collatz Conjecture | File mini_collatz_2384206285789510509683_103079215104 exists already, skipping download
4/27/2014 7:53:05 PM | Collatz Conjecture | Starting task mini_collatz_2384206285789510509683_103079215104_0
4/27/2014 7:53:05 PM | Collatz Conjecture | [sched_op] Starting scheduler request
4/27/2014 7:53:05 PM | Collatz Conjecture | Sending scheduler request: To fetch work.
4/27/2014 7:53:05 PM | Collatz Conjecture | Requesting new tasks for ATI
4/27/2014 7:53:05 PM | Collatz Conjecture | [sched_op] CPU work request: 0.00 seconds; 0.00 devices
4/27/2014 7:53:05 PM | Collatz Conjecture | [sched_op] ATI work request: 1296.00 seconds; 0.75 devices
4/27/2014 7:53:05 PM | | [http_xfer] [ID#0] HTTP: wrote 347 bytes
4/27/2014 7:53:05 PM | | [http_xfer] [ID#1] HTTP: wrote 4133 bytes
4/27/2014 7:53:06 PM | Collatz Conjecture | Scheduler request completed: got 0 new tasks
4/27/2014 7:53:06 PM | Collatz Conjecture | [sched_op] Server version 703
4/27/2014 7:53:06 PM | Collatz Conjecture | Not sending work - last request too recent: 12 sec
4/27/2014 7:53:06 PM | Collatz Conjecture | Project requested delay of 61 seconds
4/27/2014 7:53:06 PM | Collatz Conjecture | [sched_op] Deferring communication for 00:01:01
4/27/2014 7:53:06 PM | Collatz Conjecture | [sched_op] Reason: requested by project
4/27/2014 7:53:06 PM | | [http_xfer] [ID#0] HTTP: wrote 1198 bytes
4/27/2014 7:53:06 PM | | [http_xfer] [ID#0] HTTP: wrote 1237 bytes
4/27/2014 7:53:21 PM | WUProp@Home | [checkpoint] result wu_v4_1398018388_149302_0 checkpointed
4/27/2014 7:53:30 PM | | [http_xfer] [ID#0] HTTP: wrote 1209 bytes
4/27/2014 7:53:31 PM | | [http_xfer] [ID#0] HTTP: wrote 396 bytes
4/27/2014 7:53:33 PM | | [http_xfer] [ID#0] HTTP: wrote 2296 bytes
4/27/2014 7:53:36 PM | | [http_xfer] [ID#0] HTTP: wrote 377 bytes
4/27/2014 7:53:39 PM | | [http_xfer] [ID#0] HTTP: wrote 299 bytes
4/27/2014 7:53:39 PM | | [http_xfer] [ID#0] HTTP: wrote 339 bytes
4/27/2014 7:53:42 PM | | [http_xfer] [ID#0] HTTP: wrote 1045 bytes
4/27/2014 7:53:43 PM | | [http_xfer] [ID#0] HTTP: wrote 1460 bytes
4/27/2014 7:53:43 PM | | [http_xfer] [ID#0] HTTP: wrote 515 bytes
4/27/2014 7:54:19 PM | WUProp@Home | [checkpoint] result wu_v4_1398018388_149302_0 checkpointed
4/27/2014 7:54:56 PM | Collatz Conjecture | update requested by user
4/27/2014 7:54:57 PM | Collatz Conjecture | sched RPC pending: Requested by user
4/27/2014 7:54:57 PM | Collatz Conjecture | [sched_op] Starting scheduler request
4/27/2014 7:54:57 PM | Collatz Conjecture | Sending scheduler request: Requested by user.
4/27/2014 7:54:57 PM | Collatz Conjecture | Requesting new tasks for ATI
4/27/2014 7:54:57 PM | Collatz Conjecture | [sched_op] CPU work request: 0.00 seconds; 0.00 devices
4/27/2014 7:54:57 PM | Collatz Conjecture | [sched_op] ATI work request: 1296.00 seconds; 0.75 devices
4/27/2014 7:54:58 PM | | [http_xfer] [ID#0] HTTP: wrote 1198 bytes
4/27/2014 7:54:58 PM | | [http_xfer] [ID#0] HTTP: wrote 1237 bytes
4/27/2014 7:54:58 PM | | [http_xfer] [ID#1] HTTP: wrote 4435 bytes
4/27/2014 7:54:58 PM | | [http_xfer] [ID#1] HTTP: wrote 3924 bytes
4/27/2014 7:54:58 PM | Collatz Conjecture | Scheduler request completed: got 1 new tasks
4/27/2014 7:54:58 PM | Collatz Conjecture | [sched_op] Server version 703
4/27/2014 7:54:58 PM | Collatz Conjecture | Project has no tasks available
4/27/2014 7:54:58 PM | Collatz Conjecture | Project requested delay of 61 seconds
4/27/2014 7:54:58 PM | Collatz Conjecture | [sched_op] estimated total CPU task duration: 0 seconds
4/27/2014 7:54:58 PM | Collatz Conjecture | [sched_op] estimated total ATI task duration: 10403 seconds
4/27/2014 7:54:58 PM | Collatz Conjecture | [sched_op] Deferring communication for 00:01:01
4/27/2014 7:54:58 PM | Collatz Conjecture | [sched_op] Reason: requested by project
4/27/2014 7:55:01 PM | Collatz Conjecture | Started download of solo_collatz_2384126654505977390195_824633720832
4/27/2014 7:55:01 PM | Collatz Conjecture | [file_xfer] URL: http://boinc.thesonntags.com/collatz/download/1d6/solo_collatz_2384126654505977390195_824633720832
4/27/2014 7:55:01 PM | | [http_xfer] [ID#11] HTTP: wrote 46 bytes
4/27/2014 7:55:02 PM | Collatz Conjecture | [file_xfer] http op done; retval 0 (Success)
4/27/2014 7:55:02 PM | Collatz Conjecture | [file_xfer] file transfer status 0 (Success)
4/27/2014 7:55:02 PM | Collatz Conjecture | Finished download of solo_collatz_2384126654505977390195_824633720832
4/27/2014 7:55:02 PM | Collatz Conjecture | [file_xfer] Throughput 84 bytes/sec
4/27/2014 7:55:02 PM | Collatz Conjecture | Starting task solo_collatz_2384126654505977390195_824633720832_1
4/27/2014 7:55:19 PM | WUProp@Home | [checkpoint] result wu_v4_1398018388_149302_0 checkpointed
4/27/2014 7:55:40 PM | NFS@Home | update requested by user
4/27/2014 7:55:45 PM | NFS@Home | sched RPC pending: Requested by user
4/27/2014 7:55:45 PM | NFS@Home | [sched_op] Starting scheduler request
4/27/2014 7:55:45 PM | NFS@Home | Sending scheduler request: Requested by user.
4/27/2014 7:55:45 PM | NFS@Home | Reporting 1 completed tasks
4/27/2014 7:55:45 PM | NFS@Home | Not requesting tasks: don't need
4/27/2014 7:55:45 PM | NFS@Home | [sched_op] CPU work request: 0.00 seconds; 0.00 devices
4/27/2014 7:55:45 PM | NFS@Home | [sched_op] ATI work request: 0.00 seconds; 0.00 devices
4/27/2014 7:55:45 PM | | [http_xfer] [ID#1] HTTP: wrote 3133 bytes
4/27/2014 7:55:46 PM | NFS@Home | Scheduler request completed
4/27/2014 7:55:46 PM | NFS@Home | [sched_op] Server version 703
4/27/2014 7:55:46 PM | NFS@Home | Project requested delay of 7 seconds
4/27/2014 7:55:46 PM | NFS@Home | [sched_op] handle_scheduler_reply(): got ack for task GC_4_383_32208_0
4/27/2014 7:55:46 PM | NFS@Home | [sched_op] Deferring communication for 00:00:07
4/27/2014 7:55:46 PM | NFS@Home | [sched_op] Reason: requested by project
4/27/2014 7:56:01 PM | Collatz Conjecture | [sched_op] Starting scheduler request
4/27/2014 7:56:01 PM | Collatz Conjecture | Sending scheduler request: To fetch work.
4/27/2014 7:56:01 PM | Collatz Conjecture | Requesting new tasks for ATI
4/27/2014 7:56:01 PM | Collatz Conjecture | [sched_op] CPU work request: 0.00 seconds; 0.00 devices
4/27/2014 7:56:01 PM | Collatz Conjecture | [sched_op] ATI work request: 864.00 seconds; 0.50 devices
4/27/2014 7:56:03 PM | | [http_xfer] [ID#1] HTTP: wrote 4435 bytes
4/27/2014 7:56:04 PM | | [http_xfer] [ID#1] HTTP: wrote 3924 bytes
4/27/2014 7:56:04 PM | Collatz Conjecture | Scheduler request completed: got 1 new tasks
4/27/2014 7:56:04 PM | Collatz Conjecture | [sched_op] Server version 703
4/27/2014 7:56:04 PM | Collatz Conjecture | Project has no tasks available
4/27/2014 7:56:04 PM | Collatz Conjecture | Project requested delay of 61 seconds
4/27/2014 7:56:04 PM | Collatz Conjecture | [sched_op] estimated total CPU task duration: 0 seconds
4/27/2014 7:56:04 PM | Collatz Conjecture | [sched_op] estimated total ATI task duration: 10403 seconds
4/27/2014 7:56:04 PM | Collatz Conjecture | [sched_op] Deferring communication for 00:01:01
4/27/2014 7:56:04 PM | Collatz Conjecture | [sched_op] Reason: requested by project
4/27/2014 7:56:06 PM | Collatz Conjecture | Started download of solo_collatz_2384128684721985823859_824633720832
4/27/2014 7:56:06 PM | Collatz Conjecture | [file_xfer] URL: http://boinc.thesonntags.com/collatz/download/37a/solo_collatz_2384128684721985823859_824633720832
4/27/2014 7:56:06 PM | | [http_xfer] [ID#13] HTTP: wrote 46 bytes
4/27/2014 7:56:07 PM | Collatz Conjecture | [file_xfer] http op done; retval 0 (Success)
4/27/2014 7:56:07 PM | Collatz Conjecture | [file_xfer] file transfer status 0 (Success)
4/27/2014 7:56:07 PM | Collatz Conjecture | Finished download of solo_collatz_2384128684721985823859_824633720832
4/27/2014 7:56:07 PM | Collatz Conjecture | [file_xfer] Throughput 368 bytes/sec
4/27/2014 7:56:07 PM | Collatz Conjecture | Starting task solo_collatz_2384128684721985823859_824633720832_1
4/27/2014 7:56:20 PM | WUProp@Home | [checkpoint] result wu_v4_1398018388_149302_0 checkpointed
4/27/2014 7:57:05 PM | Collatz Conjecture | [sched_op] Starting scheduler request
4/27/2014 7:57:05 PM | Collatz Conjecture | Sending scheduler request: To fetch work.
4/27/2014 7:57:05 PM | Collatz Conjecture | Requesting new tasks for ATI
4/27/2014 7:57:05 PM | Collatz Conjecture | [sched_op] CPU work request: 0.00 seconds; 0.00 devices
4/27/2014 7:57:05 PM | Collatz Conjecture | [sched_op] ATI work request: 432.00 seconds; 0.25 devices
4/27/2014 7:57:06 PM | | [http_xfer] [ID#1] HTTP: wrote 4446 bytes
4/27/2014 7:57:06 PM | | [http_xfer] [ID#1] HTTP: wrote 3932 bytes
4/27/2014 7:57:06 PM | Collatz Conjecture | Scheduler request completed: got 1 new tasks
4/27/2014 7:57:06 PM | Collatz Conjecture | [sched_op] Server version 703
4/27/2014 7:57:06 PM | Collatz Conjecture | Project has no tasks available
4/27/2014 7:57:06 PM | Collatz Conjecture | Project requested delay of 61 seconds
4/27/2014 7:57:06 PM | Collatz Conjecture | [sched_op] estimated total CPU task duration: 0 seconds
4/27/2014 7:57:06 PM | Collatz Conjecture | [sched_op] estimated total ATI task duration: 26602 seconds
4/27/2014 7:57:06 PM | Collatz Conjecture | [sched_op] Deferring communication for 00:01:01
4/27/2014 7:57:06 PM | Collatz Conjecture | [sched_op] Reason: requested by project
4/27/2014 7:57:08 PM | Collatz Conjecture | Started download of mini_collatz_2384206286098748154995_103079215104
4/27/2014 7:57:08 PM | Collatz Conjecture | [file_xfer] URL: http://boinc.thesonntags.com/collatz/download/1f0/mini_collatz_2384206286098748154995_103079215104
4/27/2014 7:57:08 PM | | [http_xfer] [ID#15] HTTP: wrote 46 bytes
4/27/2014 7:57:09 PM | Collatz Conjecture | [file_xfer] http op done; retval 0 (Success)
4/27/2014 7:57:09 PM | Collatz Conjecture | [file_xfer] file transfer status 0 (Success)
4/27/2014 7:57:09 PM | Collatz Conjecture | Finished download of mini_collatz_2384206286098748154995_103079215104
4/27/2014 7:57:09 PM | Collatz Conjecture | [file_xfer] Throughput 196 bytes/sec
4/27/2014 7:57:09 PM | Collatz Conjecture | Starting task mini_collatz_2384206286098748154995_103079215104_0
4/27/2014 7:57:19 PM | WUProp@Home | [checkpoint] result wu_v4_1398018388_149302_0 checkpointed
4/27/2014 7:58:18 PM | WUProp@Home | [checkpoint] result wu_v4_1398018388_149302_0 checkpointed
4/27/2014 7:59:18 PM | WUProp@Home | [checkpoint] result wu_v4_1398018388_149302_0 checkpointed

____________

Profile Zydor
Avatar
Send message
Joined: 19 Aug 09
Posts: 364
Credit: 840,811,292
RAC: 0
Message 19255 - Posted: 28 Apr 2014, 4:56:58 UTC
Last modified: 28 Apr 2014, 4:58:50 UTC

Standard credit levels for the Solo Collatz Conjecture v6.04 (opencl_amd_gpu) could do with a sanity check .... I *think* they are significantly down on previous credit levels.

They are currently averaging around 6,000 credits whereas before they were averaging either side of 7,000 credits.

Many times a time of 520 to 535 secs results in a credit level of 5,600 - 5,700 which is unheard of before.

If that's the new levels, fine, just seems a little odd, and low.

You can see a stack of them inside Computer ID 128929
http://boinc.thesonntags.com/collatz/results.php?hostid=128929&offset=0&show_names=0&state=4&appid=

Profile Zydor
Avatar
Send message
Joined: 19 Aug 09
Posts: 364
Credit: 840,811,292
RAC: 0
Message 19256 - Posted: 28 Apr 2014, 6:38:01 UTC
Last modified: 28 Apr 2014, 7:05:01 UTC

Another oddity that's probably worth checking out, may even be connected to the same cause as below..

Ive got a bunch of Solo_Collatz Conjecture 6.0 running on the same computer as the report below.

Currently they are predicted to run for 41hrs or 71hrs. This is not the usual clock-counter problem as they are currently 9% done and the time remaining counter is changing at one second intervals, not jumping ahead of time by leaping a bunch of seconds in one go.

It does seem to be a slow mover, two to three days to complete. PC seems ok, no evidence of a problem that side. It maybe linked to the problem below where the GPU Units seem to have slowed markedly.

They are on PC ID 128929 - I cant point you at one, as none have yet finished, so I added the screen shot below. I've checked the PC, its working fine, the oddities do appear to be originating from the apps.

(Deadline clock time shown is UTC)

Profile mfc
Send message
Joined: 17 Mar 14
Posts: 5
Credit: 38,211,224
RAC: 0
Message 19257 - Posted: 28 Apr 2014, 9:24:15 UTC

Dear Slicker:

Please remove gpu applications in Mini project,

The running time of Mini tasks on gpu will be too short
(say < 10 mins in previous screenshot)
which will increase the task initial overhead and server load.

The better configuration should be:

Micro project: for ARMs only.
Mini Project: for CPU only
Solo & Large Project: for GPU only.

Thanks for your consideration.

Profile Stephen Uitti
Send message
Joined: 13 Mar 14
Posts: 8
Credit: 571,614,673
RAC: 515,719
Message 19259 - Posted: 28 Apr 2014, 11:20:59 UTC

Machine ID 141349 is a laptop with an ATI video card. A large solo task has started, and should finish in 147 days. Machine ID 142392 has an NVIDIA GTS 250, and a large solo task takes about 3 days, and should finish tomorrow night. I see in Collatz Conjecture Preferences that i can turn off large solo tasks, but 3 days seems OK. Can this be done on a machine by machine basis?

Profile Pooh Bear 27
Avatar
Send message
Joined: 1 Aug 10
Posts: 54
Credit: 108,227,920
RAC: 0
Message 19260 - Posted: 28 Apr 2014, 12:01:18 UTC

Solo has 0 WU ready to send.

Profile Bowmore
Avatar
Send message
Joined: 3 Aug 09
Posts: 18
Credit: 501,520,052
RAC: 8,581,130
Message 19261 - Posted: 28 Apr 2014, 14:41:33 UTC

I'm getting Solo WU that are double the max length as before and all are paying 10,000 exactly. Was there a protective limit of 10k that needs to be changed with the longer WU or is this the new norm?
____________

jjwhalen
Avatar
Send message
Joined: 17 Apr 10
Posts: 20
Credit: 202,526,055
RAC: 0
Message 19262 - Posted: 28 Apr 2014, 22:17:34 UTC

I'm still only getting Mini w/u's. (I actually did receive 3 Solo tasks 1 time only.) But I see that the unsent queue of Solo is now remaining at zero, presumably due to excessively high demand. I can only guess that the server load must be very high with all these short w/u's on fast GPUs. My overclocked GTX580 runs Mini in <4 min.

I agree with the spirit of mfc's post, but I wouldn't want to see it hard-coded on the server end. Better that each user set up separate Preferences for ARM, CPU only or GPU machines as needed.

Profile Stephen Uitti
Send message
Joined: 13 Mar 14
Posts: 8
Credit: 571,614,673
RAC: 515,719
Message 19263 - Posted: 29 Apr 2014, 0:07:15 UTC - in response to Message 19257.

The better configuration should be:

Micro project: for ARMs only.
Mini Project: for CPU only
Solo & Large Project: for GPU only.


My slowest GPU, an ATI on an older laptop, takes about 12 hours on a Mini 6.04 ATI4. My faster GTS 250 should take about 14 minutes for a Mini, as it's 50x faster. And, i have a GTX 650 ti, which should be 5x faster than that, or 750x faster than the ATI. The 650 ti isn't the fastest GPU out there. So, i don't think you can simply say Minis are for CPUs only. My laptop ATI GPU is similar in speed to the laptop's CPUs - perhaps 1x to 2x faster. I'd really like control for each host. An automatic assignment should work, based on a benchmark. But the benchmark would have to be the time taken for previous Collatz work units, not the BOINC benchmarks. For example, my GTS 250 got a 78 hour large work unit, but the estimate was easily ten times that.

The way i'd do it is, for a new machine, issue a micro work unit. If it takes less than ten minutes, issue a mini work unit. If that takes less than an hour then issue a solo unit, and if that takes less than an hour, go for large units - unless the user has said they don't want them. I could see someone having a fast laptop that will be on the net, or even powered on, only sometimes, and long units aren't a good match, or some other reason. I generally like long units on my desktops.

Let's say i have a unit in progress. It's showing 3:04:00 elapsed at 45%. Converting the time to decimal time, 4 (minutes) / 60 = .066. Add the 3 hours, and you get 3.066 hours. Divide this by the percent done: 3.066 / 0.45 = 6.81 hours. That's the total time for the unit. Subtract the 3.066 it has already done, 6.81 - 3.066 = 3.75 hours remaining. These calculations use the job at hand with the CPU/GPU at hand. They've been very accurate for me. But when i compare my result with the Remaining (estimated), it might say anything from 2:30 to 7:20 instead of 3:45. I've gotten reasonable accuracy even when less than 1% of a work unit has been completed, though the more, the better.

Rick A. Sponholz
Avatar
Send message
Joined: 20 Apr 14
Posts: 6
Credit: 476,294,709
RAC: 1,541,248
Message 19264 - Posted: 29 Apr 2014, 0:46:42 UTC

Slicker,
I would love to be running the large wu's, but when BOINC assigns one to my intelgpus, the due date is too short to get it processed. When assigns one to one of my GTX690, 660Ti, or 590's cores it should be alright, but I really want to utilize my intelgpu for Collatz too. Please consider giving the large wu's more time. Thanks in advance for your consideration of my request, Col. Rick A. Sponholz
____________

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 2
Message 19267 - Posted: 29 Apr 2014, 4:05:40 UTC - in response to Message 19257.

Dear Slicker:

Please remove gpu applications in Mini project,

The running time of Mini tasks on gpu will be too short
(say < 10 mins in previous screenshot)
which will increase the task initial overhead and server load.

The better configuration should be:

Micro project: for ARMs only.
Mini Project: for CPU only
Solo & Large Project: for GPU only.

Thanks for your consideration.


Your GPUs are fast. An old nVidia 8400 GS or a GT 310 etc. will take 18 hours to run a mini WU. The HD 2400 thru HD 4xxx AMD GPUs also need smaller WUs.

If you don't want your GPUs to run mini WUs but still want them to run on CPUs, uncheck mini and then check the box to run other apps if no work is available. Then your CPUs will still get mini work since there are no solo or large CPU apps.

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 2
Message 19268 - Posted: 29 Apr 2014, 4:09:08 UTC - in response to Message 19256.

Another oddity that's probably worth checking out, may even be connected to the same cause as below..

Ive got a bunch of Solo_Collatz Conjecture 6.0 running on the same computer as the report below.

Currently they are predicted to run for 41hrs or 71hrs. This is not the usual clock-counter problem as they are currently 9% done and the time remaining counter is changing at one second intervals, not jumping ahead of time by leaping a bunch of seconds in one go.

It does seem to be a slow mover, two to three days to complete. PC seems ok, no evidence of a problem that side. It maybe linked to the problem below where the GPU Units seem to have slowed markedly.

They are on PC ID 128929 - I cant point you at one, as none have yet finished, so I added the screen shot below. I've checked the PC, its working fine, the oddities do appear to be originating from the apps.

(Deadline clock time shown is UTC)



Just because solo has been switch to non-multi-size doesn't mean that people haven't aborted or errored out other multi-size WUs which is causing them to get resent. From your post,look at the WU names. The last number is the size. You have a lot of different sized WUs which will likely screw up the estimates.

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 2
Message 19269 - Posted: 29 Apr 2014, 4:14:14 UTC - in response to Message 19264.

Slicker,
I would love to be running the large wu's, but when BOINC assigns one to my intelgpus, the due date is too short to get it processed. When assigns one to one of my GTX690, 660Ti, or 590's cores it should be alright, but I really want to utilize my intelgpu for Collatz too. Please consider giving the large wu's more time. Thanks in advance for your consideration of my request, Col. Rick A. Sponholz


7 days is plenty of time to get them done as they will finish in under a day. The problem is that BOINC is just too frakking stupid to get the estimates correct as they are based on floating math and Collatz only uses integer math. So, no matter what I set the estimate to be, it is wrong because the ratio of flops to iops is different on different devices. In other words, it is using apples per day to measure oranges per day with its overly complicated algorithm that takes several WUs before it fingures out it is stupid and corrects itself. The K.I.S.S. method would have the correct estimate in under 5 seconds, but you don't get NSF grants for K.I.S.S. type of logic. ;-)

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 2
Message 19270 - Posted: 29 Apr 2014, 4:17:25 UTC - in response to Message 19260.

Solo has 0 WU ready to send.



All the multi-size WUs that people aborted or errored out are causing the problems. There are 3K available now.

gomeyer
Send message
Joined: 1 Mar 12
Posts: 4
Credit: 62,112,081
RAC: 0
Message 19271 - Posted: 29 Apr 2014, 4:22:04 UTC
Last modified: 29 Apr 2014, 4:30:32 UTC

Slicker,
My HOME venue is set for NVidia only and SOLO only, but I haven't received any SOLO WUs on the two machines using that venue. I seem to be able to get mini but they run in only 15-20 minutes on the faster of those machines which is way too short. Am I doing something wrong here?

EDIT: D'OH. The second I posted this I got a bunch. Never Mind.

Is it possible that the SOLO check box on the SCHOOL venue is not working. The last change I made was to move them to the HOME venue from SCHOOL. I did that about 10 minutes before posting.

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 2
Message 19272 - Posted: 29 Apr 2014, 4:34:34 UTC - in response to Message 19261.

I'm getting Solo WU that are double the max length as before and all are paying 10,000 exactly. Was there a protective limit of 10k that needs to be changed with the longer WU or is this the new norm?


Thanks for catching that. You guessed the cause exactly. It should be fixed now, but keep an eye on it and let me know if it isn't. I did go back and adjust the credits for users and teams of the 529 WUs where that was the problem. I didn't bother with the host credits though.

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 2
Message 19273 - Posted: 29 Apr 2014, 5:04:27 UTC - in response to Message 19255.

Standard credit levels for the Solo Collatz Conjecture v6.04 (opencl_amd_gpu) could do with a sanity check .... I *think* they are significantly down on previous credit levels.

They are currently averaging around 6,000 credits whereas before they were averaging either side of 7,000 credits.

Many times a time of 520 to 535 secs results in a credit level of 5,600 - 5,700 which is unheard of before.

If that's the new levels, fine, just seems a little odd, and low.

You can see a stack of them inside Computer ID 128929
http://boinc.thesonntags.com/collatz/results.php?hostid=128929&offset=0&show_names=0&state=4&appid=


There are no new levels. Same apps. Same credit per total steps.

That leaves the larger numbers (not larger WU size) causing branching within the code. GPU's stream processors only work in parallel. In other words, if there are 800 stream processors and they are all calculating ax + b = c they work really fast. But, if they have to process logic which requires branching (e.g. if/then/else logic, or a while loop) then the GPU processing slows down considerably because it has to wait until all the stream processors are done branching and are back to parallel processing in unison. If a single stream need to do the "else" while all the other stream processors do the "if" it will branch and they will all sit idle waiting for the one to finish. Sure, it is milliseconds at most, but it adds up with calculating hundreds of steps for billions of numbers. Also, as the GPUs get faster, to run optimally, they need to crunch more numbers at a time which increases the odds of one of the streams branching. Branching means less utilization and therefore less work done in the same amount of time and therefore less credits. We'll have to keep an eye on this.

1 · 2 · 3 · Next
Post to thread

Message boards : News : Missing Apps


Main page · Your account · Message boards


Copyright © 2018 Jon Sonntag; All rights reserved.