Trying to attach project but account wil not verify, I get verify account -204 failed.
I know my email and password are right as I logged on here fine, via web. Can you suggest a possible reason?
Thank you for the response. I've got an older phone, it is HTC incredible, ADR6300. I did reinstall the OS, although I did not deliberately do that to try to add functionality to extend battery life or anything else. I don't suppose there is a recommended OS / kernel that I could install that is known to work with Einstein@Home?
possible solution is checking your version OS. the adr6300 was shipped with version 2.1, while some were updated at time of purchase to version 2.2. you may have to download the 2.3 update and install manually. just my best guess to solve your problem.
running BOINC on a HTC Wildfire S, especially with Einstein@home is quite hard. It take around 120 hours for 57%. So that means that it is not possible to finish the WU just in time.
I noted that the small Galaxy Fame has no issues at all running 24/7.
With the Galaxy S4 however, I had to manually limit to use only 2 cores.
With 3 active cores, surface temperatures back&front become a little too hot for my comfort (but still runs with 37deg Celsius set max. battery temp).
At 4 cores full load it clearly overheats, BOINC is very frequently halting calculations for battery to cool down (touching it feels VERY unhealthy for the device then).
The S4 uses AMOLED display, which lifespan is apparently limited by temperature exposure by design (an AMOLED specific issue I've read), thus I feared for the device longlivety under these harsh conditions.
Running on default settings (use all cores, upto 40deg Celsius Battery Temp) on the Galaxy S4 therefor seemed a very bad idea.
Judging from the temperatures generated on the device surface and backside in vicinity of the CPU (far beyond 40deg Celsius, definitely uncomfortably hot for such an expensive gadget), I can only assume the battery temperature sensors are placed in a less-than-optimal spot to judge the overall device safety for long-running CPU intensive operations.
I have little doubt the CPU is auto-throttling all the time under these conditions to prevent overheating, the Display brightness setting issues the warning that the max. brightness is reduced as well (this is to prevent overheating the sensitive AMOLED). Both are signs to me that continued operation in these temperature regimes is a bad idea.
-------------
So in short, the Android Client runs
+ very stable, but
- operating temperatures on some high-end devices may easily exceed healthy regions (AMOLED displays, have no details about the effect on other components) under full load
- even with the default safety temperature set more conservative and functioning due to specific location of battery/temperature sensors (IMHO)
Not so much of an issue for experienced / techy crunchers who can judge the the effects of sustained, high temperature and who know exactly what they're doing.
Could become an issue for inexperienced volunteers who simply lack the tech knowledge, though... Wouldn't want to see high-end devices going RMA due to BOINC usage, but I fear this is a real possibility.
Right now I would rather recommend to set max. CPUs used by default to 2 or even 1 after App Installation - with a notification (and needed warnings) about the option to use more cores with respective caution towards device temperature. At least that's what feels like the much safer option to me after my experiences with the Galaxy S4.
It would also be great to have CPU loads dynamically adjusting to the point before the CPU goes into thermal throttling. However, I assume this option is likely blocked due to the Operating System denying an App such low-level hardware info.
I'm using the Galaxy S2 with the standard Boinc application and have downloaded two WU. I set to use one core only (could set it also to two, but doesn't help).
My problem is, that the computation first starts, every thing fine. But the CPU gets hotter and hotter and than the task is stopped with the message "battery temperature > 40°". I opened the lid and the battery itself is absolutely cool, the CPU (near the camera) is remarkably hot. The CPU cools down within a minute, but the task does not restart even it says "calculating" but there is no progress and the CPU does not get hot. I manually have to set the WU to pause and restart it that it really starts calculating. The phone was always connected to a socket power supply with 1000mA when I made these trials. My problem is that I cannot sit at the phone and restarting the tasks every 15min. I even set it (temporarly) to "calculate on battery", but it didn't help.
The last message in the event log is "task_state=EXECUTING for xxx from unsuspended", the status in the task list is "running", time is counting, but the CPU is cool and the taskmanager widget with CPU-load display shows no CPU-load at all.
My suggestion:
- please check if there is another reason why the WU is not calculating yet it says to do so.
- It should be possible to not fully load one CPU bit only with e.g. 50% to avoid heating up the CPU so much (issue with heat was mentioned before in this thread)
This my phone does not want to receive tasks. It says "Project back off" and tries to to download 3 tasks with no success. It is Samsung Galaxy Ace Duos GT S-6802. It happened with Berkeley's BOINC. While using NativeBOINC it does not want even to attach to a project, speaking of some violations. Checking Wifi on or off doesn't affect anything. NativeBOINC says "Wifi is not connected" though it is.
On my Galaxy Ace 2 everything works fine with one core running.
But when I set 2 core working, the processing stopped after 2 hours. It's written "running", but the percentage of work done is not increasing.
Only exiting and restarting BOINC make the tasks to work again.
Anyway I will not set 2 core again because the phone gets really unusable.
Here the report if can be of any help to developers:
Quote:
Client Messages
Mon Aug 26 00:29:48 GMT+02:00 2013|Einstein@Home|[task] task_state=EXECUTING for p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 from unsuspend
Mon Aug 26 00:29:48 GMT+02:00 2013|Einstein@Home|[task] task_state=EXECUTING for p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 from unsuspend
Mon Aug 26 00:29:48 GMT+02:00 2013||Resuming computation
Mon Aug 26 00:29:38 GMT+02:00 2013|Einstein@Home|[task] task_state=SUSPENDED for p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 from suspend
Mon Aug 26 00:29:38 GMT+02:00 2013|Einstein@Home|[task] task_state=SUSPENDED for p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 from suspend
Mon Aug 26 00:29:38 GMT+02:00 2013||Suspending computation - CPU is busy
Mon Aug 26 00:29:26 GMT+02:00 2013|Einstein@Home|Restarting task p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 using einsteinbinary_BRP4 version 106 (NEON) in slot 1
Mon Aug 26 00:29:26 GMT+02:00 2013|Einstein@Home|[task] task_state=EXECUTING for p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 from start
Mon Aug 26 00:29:26 GMT+02:00 2013|Einstein@Home|[task] ACTIVE_TASK::start(): forked process: pid 2677
Mon Aug 26 00:29:26 GMT+02:00 2013|Einstein@Home|[task] task_state=UNINITIALIZED for p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 from handle_exited_app
Mon Aug 26 00:29:26 GMT+02:00 2013|Einstein@Home|[task] process got signal 9
Mon Aug 26 00:29:26 GMT+02:00 2013|Einstein@Home|[task] Process for p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 exited, status 9, task state 1
Mon Aug 26 00:29:25 GMT+02:00 2013|Einstein@Home|Restarting task p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 using einsteinbinary_BRP4 version 106 (NEON) in slot 0
Mon Aug 26 00:29:25 GMT+02:00 2013|Einstein@Home|[task] task_state=EXECUTING for p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 from start
Mon Aug 26 00:29:25 GMT+02:00 2013|Einstein@Home|[task] ACTIVE_TASK::start(): forked process: pid 2675
Mon Aug 26 00:29:25 GMT+02:00 2013|Einstein@Home|Restarting p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 - message timeout
Mon Aug 26 00:29:25 GMT+02:00 2013|Einstein@Home|[task] task_state=UNINITIALIZED for p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 from handle_exited_app
Mon Aug 26 00:29:25 GMT+02:00 2013|Einstein@Home|[task] process got signal 9
Mon Aug 26 00:29:25 GMT+02:00 2013|Einstein@Home|[task] Process for p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 exited, status 9, task state 1
Mon Aug 26 00:29:24 GMT+02:00 2013|Einstein@Home|Restarting p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 - message timeout
Mon Aug 26 00:29:24 GMT+02:00 2013|Einstein@Home|Restarting p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 - message timeout
Mon Aug 26 00:26:33 GMT+02:00 2013|Einstein@Home|[task] task_state=EXECUTING for p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 from unsuspend
Mon Aug 26 00:26:33 GMT+02:00 2013|Einstein@Home|[task] task_state=EXECUTING for p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 from unsuspend
Mon Aug 26 00:26:33 GMT+02:00 2013||Resuming computation
Mon Aug 26 00:26:23 GMT+02:00 2013|Einstein@Home|[task] task_state=SUSPENDED for p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 from suspend
Mon Aug 26 00:26:23 GMT+02:00 2013|Einstein@Home|[task] task_state=SUSPENDED for p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 from suspend
Mon Aug 26 00:26:23 GMT+02:00 2013||Suspending computation - CPU is busy
Mon Aug 26 00:19:27 GMT+02:00 2013|Einstein@Home|[task] task_state=EXECUTING for p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 from unsuspend
Mon Aug 26 00:19:27 GMT+02:00 2013|Einstein@Home|[task] task_state=EXECUTING for p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 from unsuspend
Mon Aug 26 00:19:27 GMT+02:00 2013||Resuming computation
Mon Aug 26 00:19:16 GMT+02:00 2013|Einstein@Home|[task] task_state=SUSPENDED for p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 from suspend
Mon Aug 26 00:19:16 GMT+02:00 2013|Einstein@Home|[task] task_state=SUSPENDED for p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 from suspend
Mon Aug 26 00:19:16 GMT+02:00 2013||Suspending computation - CPU is busy
Mon Aug 26 00:18:51 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:18:22 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:17:32 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:17:12 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:16:23 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:16:03 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:15:12 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:14:53 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:14:02 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:13:42 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:12:50 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:12:31 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:11:40 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:11:20 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:10:29 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:10:11 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:09:19 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:09:01 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:08:08 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:07:36 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:06:58 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:06:27 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:05:47 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:05:21 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:04:39 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:04:16 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:03:25 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:03:08 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:02:07 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:01:49 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Mon Aug 26 00:01:02 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Mon Aug 26 00:00:33 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Sun Aug 25 23:59:52 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Sun Aug 25 23:59:21 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Sun Aug 25 23:58:38 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Sun Aug 25 23:58:09 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Sun Aug 25 23:57:29 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.N.b0s0g0.00000_1010_0 checkpointed
Sun Aug 25 23:57:00 GMT+02:00 2013|Einstein@Home|[task] result p2030.20110128.G193.95+00.25.S.b3s0g0.00000_2086_0 checkpointed
Sun Aug 25 23:56:17 GM
I crunched 4 tasks on my galaxy S1 mini @ 600 Mhz (VFP) in the last two weeks and all are validated until now.
The runtime is around 84 to 85 hours for each task.
Wifi is disabled while crunching and the device is in Airplainmode.I think that gives it a better computation time and especially a little bit more free ram
(This device has only 280mb ram and around 230 are in use while crunching.
I only enable wifi to fetch new work and report to the sheduler.
The only thing what could bettered i think is if possible the device could get new work and report the result in one sheduler contact.
In other words the device gets contact with your servers and loads new work but then i have to wait one minute and update manually to report the complete uploaded task.
That´s the only thing that disturbs me a little bit.
Or did i understand something technical wrong with the server communication mechanics?
For such a slow device it´s running really good and fast i would say.
... if possible the device could get new work and report the result in one sheduler contact.
You should be able to do this quite easily. Imagine you are at the point where you wish to upload (and report) completed tasks and also get new tasks. Before you re-enable WiFi, set your work cache size really low, below the size of the work you have left from the previous download. Then when WiFi is turned on, the completed work will upload but no new work will be requested. When the uploads complete, set your cache size back to normal and BOINC will report completed tasks and request new tasks simultaneously.
Trying to attach project but
)
Trying to attach project but account wil not verify, I get verify account -204 failed.
I know my email and password are right as I logged on here fine, via web. Can you suggest a possible reason?
Thanks
Thank you for the response.
)
Thank you for the response. I've got an older phone, it is HTC incredible, ADR6300. I did reinstall the OS, although I did not deliberately do that to try to add functionality to extend battery life or anything else. I don't suppose there is a recommended OS / kernel that I could install that is known to work with Einstein@Home?
Thanks,
Dave
possible solution is checking
)
possible solution is checking your version OS. the adr6300 was shipped with version 2.1, while some were updated at time of purchase to version 2.2. you may have to download the 2.3 update and install manually. just my best guess to solve your problem.
hello, running BOINC on a
)
hello,
running BOINC on a HTC Wildfire S, especially with Einstein@home is quite hard. It take around 120 hours for 57%. So that means that it is not possible to finish the WU just in time.
philipp
Been running Android on two
)
Been running Android on two devices since a few weeks.
Samsung Galaxy Fame (1GHz SingleCore)
Samsung Galaxy S4 (1.9GHz Quadcore, European Version)
I noted that the small Galaxy Fame has no issues at all running 24/7.
With the Galaxy S4 however, I had to manually limit to use only 2 cores.
With 3 active cores, surface temperatures back&front become a little too hot for my comfort (but still runs with 37deg Celsius set max. battery temp).
At 4 cores full load it clearly overheats, BOINC is very frequently halting calculations for battery to cool down (touching it feels VERY unhealthy for the device then).
The S4 uses AMOLED display, which lifespan is apparently limited by temperature exposure by design (an AMOLED specific issue I've read), thus I feared for the device longlivety under these harsh conditions.
Running on default settings (use all cores, upto 40deg Celsius Battery Temp) on the Galaxy S4 therefor seemed a very bad idea.
Judging from the temperatures generated on the device surface and backside in vicinity of the CPU (far beyond 40deg Celsius, definitely uncomfortably hot for such an expensive gadget), I can only assume the battery temperature sensors are placed in a less-than-optimal spot to judge the overall device safety for long-running CPU intensive operations.
I have little doubt the CPU is auto-throttling all the time under these conditions to prevent overheating, the Display brightness setting issues the warning that the max. brightness is reduced as well (this is to prevent overheating the sensitive AMOLED). Both are signs to me that continued operation in these temperature regimes is a bad idea.
-------------
So in short, the Android Client runs
+ very stable, but
- operating temperatures on some high-end devices may easily exceed healthy regions (AMOLED displays, have no details about the effect on other components) under full load
- even with the default safety temperature set more conservative and functioning due to specific location of battery/temperature sensors (IMHO)
Not so much of an issue for experienced / techy crunchers who can judge the the effects of sustained, high temperature and who know exactly what they're doing.
Could become an issue for inexperienced volunteers who simply lack the tech knowledge, though... Wouldn't want to see high-end devices going RMA due to BOINC usage, but I fear this is a real possibility.
Right now I would rather recommend to set max. CPUs used by default to 2 or even 1 after App Installation - with a notification (and needed warnings) about the option to use more cores with respective caution towards device temperature. At least that's what feels like the much safer option to me after my experiences with the Galaxy S4.
It would also be great to have CPU loads dynamically adjusting to the point before the CPU goes into thermal throttling. However, I assume this option is likely blocked due to the Operating System denying an App such low-level hardware info.
I'm using the Galaxy S2 with
)
I'm using the Galaxy S2 with the standard Boinc application and have downloaded two WU. I set to use one core only (could set it also to two, but doesn't help).
My problem is, that the computation first starts, every thing fine. But the CPU gets hotter and hotter and than the task is stopped with the message "battery temperature > 40°". I opened the lid and the battery itself is absolutely cool, the CPU (near the camera) is remarkably hot. The CPU cools down within a minute, but the task does not restart even it says "calculating" but there is no progress and the CPU does not get hot. I manually have to set the WU to pause and restart it that it really starts calculating. The phone was always connected to a socket power supply with 1000mA when I made these trials. My problem is that I cannot sit at the phone and restarting the tasks every 15min. I even set it (temporarly) to "calculate on battery", but it didn't help.
The last message in the event log is "task_state=EXECUTING for xxx from unsuspended", the status in the task list is "running", time is counting, but the CPU is cool and the taskmanager widget with CPU-load display shows no CPU-load at all.
My suggestion:
- please check if there is another reason why the WU is not calculating yet it says to do so.
- It should be possible to not fully load one CPU bit only with e.g. 50% to avoid heating up the CPU so much (issue with heat was mentioned before in this thread)
This my phone does not want
)
This my phone does not want to receive tasks. It says "Project back off" and tries to to download 3 tasks with no success. It is Samsung Galaxy Ace Duos GT S-6802. It happened with Berkeley's BOINC. While using NativeBOINC it does not want even to attach to a project, speaking of some violations. Checking Wifi on or off doesn't affect anything. NativeBOINC says "Wifi is not connected" though it is.
On my Galaxy Ace 2 everything
)
On my Galaxy Ace 2 everything works fine with one core running.
But when I set 2 core working, the processing stopped after 2 hours. It's written "running", but the percentage of work done is not increasing.
Only exiting and restarting BOINC make the tasks to work again.
Anyway I will not set 2 core again because the phone gets really unusable.
Here the report if can be of any help to developers:
I crunched 4 tasks on my
)
I crunched 4 tasks on my galaxy S1 mini @ 600 Mhz (VFP) in the last two weeks and all are validated until now.
The runtime is around 84 to 85 hours for each task.
Wifi is disabled while crunching and the device is in Airplainmode.I think that gives it a better computation time and especially a little bit more free ram
(This device has only 280mb ram and around 230 are in use while crunching.
I only enable wifi to fetch new work and report to the sheduler.
The only thing what could bettered i think is if possible the device could get new work and report the result in one sheduler contact.
In other words the device gets contact with your servers and loads new work but then i have to wait one minute and update manually to report the complete uploaded task.
That´s the only thing that disturbs me a little bit.
Or did i understand something technical wrong with the server communication mechanics?
For such a slow device it´s running really good and fast i would say.
RE: ... if possible the
)
You should be able to do this quite easily. Imagine you are at the point where you wish to upload (and report) completed tasks and also get new tasks. Before you re-enable WiFi, set your work cache size really low, below the size of the work you have left from the previous download. Then when WiFi is turned on, the completed work will upload but no new work will be requested. When the uploads complete, set your cache size back to normal and BOINC will report completed tasks and request new tasks simultaneously.
Is this what you are trying to do?
Cheers,
Gary.