3 day deadlines?
At least WCG gives one 7-10 days. :-)

I reduced cache to 0.1/0 and runtime to minimal of 2h. Will see how this would help.
But not sure I understand how all this could help with cited runtimes.

It will (eventually) help once Estimated times match reality and it uses the Fraction of time BOINC is running xxx% and While BOINC is running, fraction of time computing is allowed xxx% values to set the cache size.
And with the shorter runtime there would be less stops & restarts for a given Task- from memory it can be 5 or so minutes between Checkpoints, and that's on my i7-8700K. I've no idea what the biggest gap between checkpoints on a phone might be.
But with such a big difference between CPU time & Runtime even that may not be enough.

Elapsed and CPU times depend exclusively from phone usage _and_ application (can it use CPU all time or waste much time on waiting for disk/memory), not from cache size or Rosetta's projected run time.

Yeah.
Rosetta has a lot of disk activity at certain points during a Task.... and when starting & finishing a Task- so starting & stopping computation, frequently, really works everything hard. And with limited RAM (and even with lots of RAM in some cases) Leave in memory while suspended just is not an option.

So the only thing that comes to mind is the "Compute only between ---" option. It'll reduce the difference between CPU time & Run time, and end all the stopping & starting of the Tasks.


Even if Rosetta were to explicitly state their applications are not suitable for Mobile phones, people would still use them & complain. Or complain that they don't have an application they can use on their phone...

---

Grant
Darwin NT

It seems Rosetta isn't suitable for Android smartphones.
Even for those who can deal with its disk space and memory requirements.
https://boinc.bakerlab.org/rosetta/results.php?hostid=4069076
Too much time/energy is lost for nothing.

You can select how long it runs for, you can select the cache size, you can set when it does or doesn't process work so you don't end up with things like this.

Run time 1 hours 10 min 24 sec
CPU time 24 min 41 sec

So, if you want to have long runtimes and a huge cache, and the phone doesn't spend much of it's time actually processing BOINC work then most definitely- Rosetta is not the project for you.

I reduced cache to 0.1/0 and runtime to minimal of 2h. Will see how this would help.
But not sure I understand how all this could help with cited runtimes.
Elapsed and CPU times depend exclusively from phone usage _and_ application (can it use CPU all time or waste much time on waiting for disk/memory), not from cache size or Rosetta's projected run time.

and does the latest version of the Android BOINC Manager have the fix for this in it?

<message>
finish file present too long
</message>

?

BOINC was installed on that phone week or two ago. So perhaps it has latest version. But I'll check for update.

EDIT: the reason of exit with error code at least for 2 tasks was "too much restarts". BOINC on Android can't prevent restarts by keeping stuff in memory. Just because BOINC get killed by itself from memory by OS. So if phone in active use restarts are inevitable.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

It seems Rosetta isn't suitable for Android smartphones.
Even for those who can deal with its disk space and memory requirements.
https://boinc.bakerlab.org/rosetta/results.php?hostid=4069076
Too much time/energy is lost for nothing.

You can select how long it runs for, you can select the cache size, you can set when it does or doesn't process work so you don't end up with things like this.

Run time 1 hours 10 min 24 sec
CPU time 24 min 41 sec

So, if you want to have long runtimes and a huge cache, and the phone doesn't spend much of it's time actually processing BOINC work then most definitely- Rosetta is not the project for you.

and does the latest version of the Android BOINC Manager have the fix for this in it?

<message>
finish file present too long
</message>

?

---

Grant
Darwin NT

It seems Rosetta isn't suitable for Android smartphones.
Even for those who can deal with its disk space and memory requirements.
https://boinc.bakerlab.org/rosetta/results.php?hostid=4069076
Too much time/energy is lost for nothing.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Congrats!
Join us on Rosetta too for a while :)

Sorry, my CPUs are needed to run the GPUs. I'm not about to turn off 8 GPUs to run 8 CPU tasks. Now, if Rosetta can produce a GPU App that works similar to the SETI or Einstein NV Apps I'll give it some time.

BTW, the main GPU Host is up to #16 now, and should have clear sailing up to around #4.

Time for a MSI B360F-Pro MB and cpu (you can get an 8c/16t one) upgrade!



ROFLing.....


Sorry I could not resist. Actually if its not broke don't "fix it" !!!


Tom M

---

"I owe", "I owe", "Its off to work I go" (from a bumper sticker on a smallish Mercedes Benz)
(on the back of a Semi Tractor) "If you can read this bumper sticker, I've LOST MY TRAILER!"

Don't forget to join WCG too:

My Project and Team Recruitment Link

Congrats!
Join us on Rosetta too for a while :)

Sorry, my CPUs are needed to run the GPUs. I'm not about to turn off 8 GPUs to run 8 CPU tasks. Now, if Rosetta can produce a GPU App that works similar to the SETI or Einstein NV Apps I'll give it some time.

BTW, the main GPU Host is up to #16 now, and should have clear sailing up to around #4.

If I wasn't already in the GPU Usersgroup I would join. Sorry.

Tom

Same holds me on E@h. Already in Moon's SICITURADASTRA team.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

So.... look what can happen in a mere 1 week 4 days. I went from Zero to 11th place at Einstein, https://einsteinathome.org/community/stats/users.
Of course, things are still changing, and yesterday's numbers show an output of over 7 Million, so, it will probably go a little higher...especially if I get some of that Helicopter money from D.C.

Congrats!
Join us on Rosetta too for a while :)

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

So.... look what can happen in a mere 1 week 4 days. I went from Zero to 11th place at Einstein, https://einsteinathome.org/community/stats/users.
Of course, things are still changing, and yesterday's numbers show an output of over 7 Million, so, it will probably go a little higher...especially if I get some of that Helicopter money from D.C.

And 18th place for a system.

I am pumping up just under 400,000 RAC on a 2 gpu Nvidia system (and loath to change to Amd/Ati from my Nvidia's). So it sure looks like any gpu-heavy system should be VERY productive.

Congratulations Tbar!

Tom

---

"I owe", "I owe", "Its off to work I go" (from a bumper sticker on a smallish Mercedes Benz)
(on the back of a Semi Tractor) "If you can read this bumper sticker, I've LOST MY TRAILER!"

We have 2 new members of Seti Orphans at Rosetta

Boiler Paul and BettyA

Welcome

Come on guys, let's make it BIG.

+1

If I wasn't already in the GPU Usersgroup I would join. Sorry.

Tom

---

"I owe", "I owe", "Its off to work I go" (from a bumper sticker on a smallish Mercedes Benz)
(on the back of a Semi Tractor) "If you can read this bumper sticker, I've LOST MY TRAILER!"

We have 2 new members of Seti Orphans at Rosetta

Boiler Paul and BettyA

Welcome

Come on guys, let's make it BIG.

---

Dave

So.... look what can happen in a mere 1 week 4 days. I went from Zero to 11th place at Einstein, https://einsteinathome.org/community/stats/users.
Of course, things are still changing, and yesterday's numbers show an output of over 7 Million, so, it will probably go a little higher...especially if I get some of that Helicopter money from D.C.

And 18th place for a system.

---

Grant
Darwin NT

So.... look what can happen in a mere 1 week 4 days. I went from Zero to 11th place at Einstein, https://einsteinathome.org/community/stats/users.
Of course, things are still changing, and yesterday's numbers show an output of over 7 Million, so, it will probably go a little higher...especially if I get some of that Helicopter money from D.C.

yes thanks for the Rosetta help posts. I've shortened my tasks from 8 to 4 hours because pausing them when I get seti tasks was causing errors. With only 4 hours per task I can let the Rosetta ones finish before starting seti tasks (not that I get many- but I like prioritizing them).

I'm also asking for Rosetta in small batches (1 day) and then NNT the project. It is working and gives me something to play with while I'm stuck at home.

Thank you all a lot for answers, they definitely improved my understanding of Rosetta workflow. Will be usefull for other "refugees " too IMHO.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

If you move the allowed task computation time while the cache has tasks you can end up with a lot of "runs longer than it should" tasks. My understanding is the "Watch Dog" timer will kick if the task runs more than 4 hours past the "current" allowed task computation.

Nope, the "watch dog" gets the info about the new target runtime, did that myself, no issues at all.

---

Don't know about the others but, Rosetta has work.,

---

Dave

Also got my billion (Boinc) with a little help. FAH, Rosetta, GPUGRID are all short of work now, must be due to us lot!

---

"none so blind as those who will not see" (John Heywood 1546)

Don't drink water, that stuff rusts pipes!


You are making Proof out of Logic, by just being dubious! {Bluestar to me)