| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
| World Community Grid Forums
|
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 9
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hey all, I hope everyone is doing fine!
I've been running BOINC on multiple devices and out of them 4 are Android phones. With the new Work Units from Open Pandemics - COVID 19, I've observed that they start pretty quickly, but reset to 0 on their own after 4-5%. The most recent was 4 WUs which reset to 0 after 4.2%. Time taken to reach that level is about 8-10 minutes. I've observed similar thing on all phones. And once that happens, the processing becomes really slow. Less than 1% in even 20 minutes! Has anyone else also faced such a problem? I'm sorry if this is not the correct place to post this. I'd love for someone to point me in the right direction. |
||
|
|
yoerik
Senior Cruncher Canada Joined: Mar 24, 2020 Post Count: 413 Status: Offline Project Badges:
|
I'm no expert - but I had a similar issue with MCM and Rosetta@home on my Android devices.
----------------------------------------For the resetting to 0% issue - How much RAM do your devices have, and how many OP WUs are running at once? Each OP WU requires 250 MB of RAM accessible each. Not having access to the necessary RAM when starting, I've found causes that error - starting and resetting, etc. Once they get going, they need less RAM overall. However - my theory would be that 250 MB is needed to launch and usage fluctuates as it crunches... so if you're trying to launch several at once, that causes the error. Could be completely wrong. For slow progression - I wouldn't call that a problem per say. Mobile CPUs are significantly weaker than desktop/laptop CPUs. My OP WUs are running over 20 hours, sometimes 24 hours per WU. It depends on how many WUs are running and how I'm using my devices. ![]() [Edit 1 times, last edit by yoerik at May 16, 2020 9:29:55 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Thanks for your reply.
One of the devices has 6GB RAM. I don't play any games or run multiple apps. I have also locked the BOINC app so that it gets neither killed in the background not gets suspended due to no activity. And there are 4 WUs running at a time. The phone is a Huawei Honor 8 Pro - 6 GB RAM, 128 GB storage, 8-core processor with Max frequency of 2.36 GHz. Although that is seldom reached because Android mostly utilizes the little core cluster to save battery (even when on charge) and I've seen using CPU-Z, the 4-cores stay at 1844 MHz pretty much always. And the rest 4 at 933 MHz and jump to 2360 MHz as and when need be. Does this provide any helpful information? |
||
|
|
yoerik
Senior Cruncher Canada Joined: Mar 24, 2020 Post Count: 413 Status: Offline Project Badges:
|
Thanks for your reply. One of the devices has 6GB RAM. I don't play any games or run multiple apps. I have also locked the BOINC app so that it gets neither killed in the background not gets suspended due to no activity. And there are 4 WUs running at a time. The phone is a Huawei Honor 8 Pro - 6 GB RAM, 128 GB storage, 8-core processor with Max frequency of 2.36 GHz. Although that is seldom reached because Android mostly utilizes the little core cluster to save battery (even when on charge) and I've seen using CPU-Z, the 4-cores stay at 1844 MHz pretty much always. And the rest 4 at 933 MHz and jump to 2360 MHz as and when need be. Does this provide any helpful information? well the Huawei shouldn't be having issues with RAM, 250 MB*8 cores = maximum need of 2 GB of RAM. Again, I'm no expert - I can't help interpret the MHz's potential impact on your WU runtimes (but hopefully a more experienced cruncher will come along and be able to help ;p) Basically though - your Huawei device with that 6 GB RAM shouldn't be running into that resetting to 0% issue. If it is, there's another issue - either with the device or the WUs. ![]() |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
In the BOINC app go to preferences, tick "Show advanced preferences..." then scroll down and verify enough storage/memory is allowed. At bottom set "Boinc client log flags" and tick android_debug, task_debug, gui_rpc_debug and save. Then let it run till you see another reset. When it has happened, go to event log, hit the copy icon and post the log.
----------------------------------------Notably, minimum Android version is 4.4.2. Edit: For good order, the app is looking for 'best/lowest docking energy" possible and that is retried until a certain minimum threshold is met. I'm guessing that it's doing that for the first docking pass and then retreats after which the total estimated runtime is re calibrated. [Edit 1 times, last edit by Former Member at May 16, 2020 10:55:33 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The settings earlier were to allow using 80% of the storage space and 70% RAM. I changed it to 100% for both this time.
Enabled the debug flags and started 2 new WUs. This time the reset happened even before 2% I think. The logs are here : https://pastebin.com/zxTb8vsS |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Nothing I see quick parsing the log indicating bad operation and you've not said the tasks croak, so think you're just being confused by that alternating progress looking for that lowest energy more explained here https://www.worldcommunitygrid.org/forums/wcg...read,11332_offset,0#84461
The present graphics efforts by WCG compared to past are not their best work, but this is how part of old AutoDock graphics looked here. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Yeah, I get it now.
Many thanks for the help and sharing the link, it indeed was very helpful and cleared the confusion Happy crunching! |
||
|
|
|