Please post issues regarding Rosetta and Rosetta android 4.1+ application versions here.

---

1st issue

On one device there's download the files, but won't crunsh anything. Says "download error".

Android 7.1.x 64-bit Amlogic

On Android 6.0 Marshmellow NO ISSUES, it crunshes well.

Hmmm, any other information? Let us know if it persists.

I had a whole batch of WUs error with 30 seconds runtime and 10 seconds CPU time. The device does not normally have any issues (has over 100k credits).

Rosetta for Android v4.10 running on Android 6.0.1

<core_client_version>7.4.53</core_client_version>
<![CDATA[
<message>
process exited with code 1 (0x1, -255)
</message>
<stderr_txt>
command: ../../projects/boinc.bakerlab.org_rosetta/rosetta_android_4.10_arm-android-linux-gnu @cispro_backbone_scaffold_test_7res_PPXXXXX.flags -nstruct 10000 -cpu_run_time 14400 -watchdog -boinc:max_nstruct 600 -checkpoint_interval 120 -mute all -database minirosetta_database -in::file::zip minirosetta_database.zip -boinc::watchdog -run::rng mt19937 -constant_seed -jran 3911373

ERROR: in::file::boinc_wu_zip cispro_backbone_scaffold_7res_PPXXXXX.zip does not exist!
ERROR:: Exit from: src/apps/public/boinc/minirosetta.cc line: 180
BOINC:: Error reading and gzipping output datafile: default.out
called boinc_finish(1)

</stderr_txt>
]]>

Application version Rosetta for Android v4.10
Device 3396190, Task 998505726, and WU 899502854.
Status: Error while computing.
Errors: Too many errors (may have bug). Too many total results.
Exit status: 1 (0x00000001) Unknown error code

CPU type ARM AArch64 Processor rev 3 (aarch64)
Operating System Android 3.18.19-g407ac4f82ff (Android 5.1.1)
Device: Amazon Fire tablet

<message> process exited with code 1 (0x1, -255) </message>
WARNING: linker: ../../projects/boinc.bakerlab.org_rosetta/rosetta_android_4.10_arm-android-linux-gnu: unused DT entry: type 0x6ffffffe arg 0x26cc
WARNING: linker: ../../projects/boinc.bakerlab.org_rosetta/rosetta_android_4.10_arm-android-linux-gnu: unused DT entry: type 0x6fffffff arg 0x2

ERROR: in::file::boinc_wu_zip cispro_backbone_scaffold_7res_PPXXXXX.zip does not exist!
ERROR:: Exit from: src/apps/public/boinc/minirosetta.cc line: 180
BOINC:: Error reading and gzipping output datafile: default.out called boinc_finish(1)

Note that mine was 2nd host to try to process this WU, with basically the same error other than the WARNINGs above.

Also had same error with 2 other WUs:
899502576
899503341

I think it's not as light as it should be, I set BOINC to run 100% of CPUs= 1=8 Cores, and 100% of the time, the processes run on Low Priority on Win 8.1 64-bit, btu they hog the computer sometimes, not as responsive as it used to be, so please fix this, same thing on the Android client, if you make it really transparent as it used to be a few months/years ago would be better for everyone, because sometimes I need to pause the computation and I forget about it, and BOINC resumes after 1 hour later, so I'd like to avoid to pause BOINC

Double post

No errors, but I've stopped receiving tasks.

I finished out a queue overnight and I'm not getting any new ones from the server. Running on Android.

There are currently no tasks available for Android.
See server status page...

---

Why mine when you can research? - GRIDCOIN - Real cryptocurrency without wasting hashes! www.gridcoin.us

What has happened to Android WU ?
Nothing for over a week now.
No WU unsent, very few in progress.

Look here: https://boinc.bakerlab.org/rosetta/server_status.php

There are no work units available for android it seems. 204 units running with 13 users in total. While there are around 18 000 users running PC and Mac

---

Look here: https://boinc.bakerlab.org/rosetta/server_status.php

There are no work units available for android it seems. 204 units running with 13 users in total. While there are around 18 000 users running PC and Mac

Right now, there are
0 Rosetta Android 4.0 jobs in queue for 204 users and
3 Rosetta 4.0 jobs in the queue for 8192 PC/MAC users (I am not sure what the average number of computers per "user" is).
14,985 Rosetta 3.78 jobs in the queue.

Android cannot run the 3.78 jobs. My guess is that the researchers can select either 3.78 or 4.0 applications and many select the 3.78 because they trust their jobs will complete. There has been a regular failure of the 4.0 jobs on the newer Linux distributions.

Once David gets the new Rosetta with the putenv() fix qualified, I suspect that there will be less researcher focus on running Rosetta 3.78 jobs. I don't know why the server needed to distinguish between any of the 4.0 application computers and should be able to just deploy to the next requesting machine.

The Rosetta job dispatcher, however, does not seem to be one of the more robust or automated implementations.

The server Status page finally shows Android tasks - 4655 right now

I currently can't get any because my buffer is full of WCG tasks, but they should start coming through very soon, hopefully

---

The server Status page finally shows Android tasks - 4655 right now

I currently can't get any because my buffer is full of WCG tasks, but they should start coming through very soon, hopefully

When I want to fix that, I just SUSPEND the project (WCG in this case) for a few seconds and disrupt the scheduler. It downloads the new WU and then rotates between them. BUT, .... I have my buffer set at 0.1 day's work AND several backup projects at 0 resource.

The Server Status page finally shows Android tasks - 4655 right now

I currently can't get any because my buffer is full of WCG tasks, but they should start coming through very soon, hopefully

When I want to fix that, I just SUSPEND the project (WCG in this case) for a few seconds and disrupt the scheduler. It downloads the new WU and then rotates between them. BUT... I have my buffer set at 0.1 day's work AND several backup projects at 0 resource.

I couldn't be bothered at the time, but I've set WCG to No New Tasks now.

And, of course, all Android tasks have gone to others and no more have come through, so this situation could linger further. Everyone stand down...

Ah well, I'll take another glance in the morning.

---

Tell a lie, 4 came down and I already completed them during the day, only to find none further to grab. Ok, that's not quite so bad. At least I got my share while they were available.

---

Finally took a look at my android tasks - still none, but a quick update brought 4 down and Server Status shows ~10k to grab.

Could we be back on stream?

---

Back to regular flow. :)

---

Application version Rosetta for Android v4.10
Device 3396190, Task 1059803643, and WU 954701287.
Status: Error while computing.
Errors: Too many errors (may have bug). Too many total results.
Exit status: 0 (0x00000000)

CPU type ARM AArch64 Processor rev 3 (aarch64)
Operating System Android 3.18.19-g407ac4f82ff (Android 5.1.1)
Device: Amazon Fire tablet

WARNING: linker: ../../projects/boinc.bakerlab.org_rosetta/rosetta_android_4.10_arm-android-linux-gnu: unused DT entry: type 0x6ffffffe arg 0x26cc
WARNING: linker: ../../projects/boinc.bakerlab.org_rosetta/rosetta_android_4.10_arm-android-linux-gnu: unused DT entry: type 0x6fffffff arg 0x2
command: ../../projects/boinc.bakerlab.org_rosetta/rosetta_android_4.10_arm-android-linux-gnu @rb_02_25_1332_1477_ab_t000__robetta_FLAGS -in::file::fasta t000_.fasta -psipred_ss2 t000_.spider3_ss2 -kill_hairpins t000_.nobuformat.spider3_ss2 -abinitio::use_filters true -constraints::cst_file t000_.fasta.CB.cst -constraints:cst_weight 5.0 -constraints::cst_fa_file t000_.fasta.MIN.cst -constraints:cst_fa_weight 5.0 -in:file:boinc_wu_zip rb_02_25_1332_1477_ab_t000__robetta.zip -frag3 rb_02_25_1332_1477_ab_t000__robetta.200.3mers.index.gz -fragA rb_02_25_1332_1477_ab_t000__robetta.200.15mers.index.gz -fragB rb_02_25_1332_1477_ab_t000__robetta.200.4mers.index.gz -nstruct 10000 -cpu_run_time 14400 -watchdog -boinc:max_nstruct 600 -checkpoint_interval 120 -mute all -database minirosetta_database -in::file::zip minirosetta_database.zip -boinc::watchdog -run::rng mt19937 -constant_seed -jran 1984289

Too many restarts with no progress. Keep application in memory while preempted.

The two warnings above repeat numerous times.

</stderr_txt>
<message>upload failure: <file_xfer_error>
<file_name>rb_02_25_1332_1477_ab_t000__robetta_cstwt_5.0_IGNORE_THE_REST_04_15_818638_16_1_r905537798_0</file_name>
<error_code>-161 (not found)</error_code>
</file_xfer_error>

Hello!

Doing compute on Samsung A5 (2016) [host 3576143] in parallel with World Community Grid (OpenZika). Long time (more than one month) compute Open Zika without any problems on all cores of Exynos 7580 (8 cores ARM Cortex A-53, 1.6 GHz, but real frequency depends of temperature) and 2 Gb RAM.

In first set of Rosetta 4.1 Android tasks from 24 items - 16 completed successfully and 8 failed at final stages of computing (run time of failed tasks near the run time of successful tasks). Response of smartphone GUI got close to 1 or even 2 seconds. After reduce of number of CPU consumed by BOINC from 8 to 7, GUI response return to small latency and next 20 tasks - completed successfully, but sometimes BOINC stops the computing. Now, with computing on 7 cores, about 260 - 300 Mb RAM is free. May be problems with tasks in first set were caused by insufficiently of RAM?

Another problem - tasks form Rosetta@Home placed in queue before any WGC tasks and occupy all available for BOINC cores (7 vs 0, instead of 3+4 or 4+3). If it had not happened, RAM consumption may be decrease and I can use all cores for computing and would not face the problem of stopping calculations.

Thank you for new application! With best wishes!

---