1131866 - MarionetteJS can't be executed on device: runner-service err: ERROR:tornado.general:Uncaught exception, closing connection

Status: RESOLVED WORKSFORME

(Testing Graveyard :: JSMarionette, defect)

Description

[Johan Lorenzo [:jlorenzo]]

This bug has been forked from bug 1091680.

STR
1. From the repo, execute: BUILDAPP=device TEST_FILES=./apps/calendar/test/marionette/day_view_test.js make test-integration
2. Wait until the execution finishes

Actual results
Nothing happens on the device at all. On the shell, you see this output: 
> ./apps/calendar/test/marionette/day_view_test.js failed. Will retry.

And after all the retries done:
> runner-service err: ERROR:tornado.general:Uncaught exception, closing connection.
> Traceback (most recent call last):
>   File "/home/mozilla/B2G/gaia/node_modules/marionette-socket-host/venv/local/lib/python2.7/site-packages/pyzmq-14.4.1-py2.7-linux-x86_64.egg/zmq/eventloop/zmqstream.py", line 407, in _run_callback
>     callback(*args, **kwargs)
>   File "/home/mozilla/B2G/gaia/node_modules/marionette-socket-host/venv/local/lib/python2.7/site-packages/tornado-4.0.2-py2.7-linux-x86_64.egg/tornado/stack_context.py", line 275, in null_wrapper
>     return fn(*args, **kwargs)
>   File "/home/mozilla/B2G/gaia/node_modules/marionette-socket-host/venv/local/lib/python2.7/site-packages/corredor-0.1-py2.7.egg/corredor/pattern.py", line 95, in on_recv
>     self.handler(data)
>   File "/home/mozilla/B2G/gaia/node_modules/marionette-socket-host/venv/local/lib/python2.7/site-packages/corredor-0.1-py2.7.egg/corredor/handler.py", line 41, in __call__
>     self.action_map[data['action']](data)
>   File "/home/mozilla/B2G/gaia/node_modules/marionette-socket-host/python/runner-service/runner_service/listener.py", line 20, in _
>     handler(data)
>   File "/home/mozilla/B2G/gaia/node_modules/marionette-socket-host/python/runner-service/runner_service/handlers/runner.py", line 128, in start_runner
>     self.runner.start()
>   File "/home/mozilla/B2G/gaia/node_modules/marionette-socket-host/venv/local/lib/python2.7/site-packages/mozrunner-6.5-py2.7.egg/mozrunner/base/device.py", line 68, in start
>     self.device.setup_profile(self.profile)
>   File "/home/mozilla/B2G/gaia/node_modules/marionette-socket-host/venv/local/lib/python2.7/site-packages/mozrunner-6.5-py2.7.egg/mozrunner/devices/base.py", line 113, in setup_profile
>     self.app_ctx.setup_profile(profile)
>   File "/home/mozilla/B2G/gaia/node_modules/marionette-socket-host/venv/local/lib/python2.7/site-packages/mozrunner-6.5-py2.7.egg/mozrunner/application.py", line 148, in setup_profile
>     self.dm.removeFile(self.remote_settings_db)
>   File "/home/mozilla/B2G/gaia/node_modules/marionette-socket-host/venv/local/lib/python2.7/site-packages/mozrunner-6.5-py2.7.egg/mozrunner/application.py", line 120, in remote_settings_db
>     raise DMError("Could not find settings db in '%s'!" % self.remote_idb_dir)
> DMError: Could not find settings db in '/data/local/storage/persistent/chrome/idb'!


Tested with, gaia revision on computer: 8c7865486a1b11076b849bbf8f7fccbaffbfafe7
Tested against a device with: 
Build ID               20150210010523
Gaia Revision          0cf517083f7eb5fc269e1236edba50534f65e3cd
Gaia Date              2015-02-09 18:21:45
Gecko Revision         https://hg.mozilla.org/mozilla-central/rev/2cb22c058add
Gecko Version          38.0a1
Device Name            flame
Firmware(Release)      4.4.2
Firmware(Incremental)  eng.cltbld.20150210.050217
Firmware Date          Tue Feb 10 05:02:28 EST 2015
Bootloader             L1TC000118D0

Comment 1

[Johan Lorenzo [:jlorenzo]]

2 different issues were filed in bug 1091680. As discussed with Aus, this one is the one where QA is blocked on. Aus told me you were 3 working on bug 1091680. Is it the case with this one?

Thanks!

Comment 2

[James Lal [:lightsofapollo]]

correct the patch gareth is working on landing entirely removes this set of dependencies which cause this error.

Comment 3

[Aus Lacroix [:aus]]

Attachment: Pull Request - Update marionette-js-runner usage and bootstrap process for tests

Comment 4

[Johan Lorenzo [:jlorenzo]]

Hi Aus! Do you have any update on this bug?

Comment 5

[Aus Lacroix [:aus]]

(In reply to Johan Lorenzo [:jlorenzo] (QA) from comment #4)
> Hi Aus! Do you have any update on this bug?

Johan, I'll be working to get this fixed next week. We're a good deal of the way there already. Basically, we need to update everyone to use a more recent version of marionette-js and runner.

Comment 6

[bhavana bajaj [:bajaj]]

aus/James any update on this yet?

Comment 7

[Aus Lacroix [:aus]]

I was working with Gareth to fix some issues he was running into last Friday to complete testing his patch. We're hoping we have a better idea of what's left after that. I'll let :gaye provide more information about when it may land.

Comment 8

[Gareth Aye [:gaye] (back from PTO)]

That patch should land later this week (perhaps even tonight) but there's still going to be a good bit of work to get marionette js tests on device to a good place.

Comment 9

[Johan Lorenzo [:jlorenzo]]

I ran BUILDAPP=device TEST_FILES=./apps/calendar/test/marionette/day_view_test.js make test-integration today and I was able to run some of the tests.

The screen often gets black but it seems like the tests are running. Anyway, this issue is no more. Let's close it.

Comment 10

[Martijn Wargers (dead)]

(In reply to Johan Lorenzo [:jlorenzo] (QA) from comment #9)
> I ran BUILDAPP=device
> TEST_FILES=./apps/calendar/test/marionette/day_view_test.js make
> test-integration today and I was able to run some of the tests.
> 
> The screen often gets black but it seems like the tests are running. Anyway,
> this issue is no more. Let's close it.

I think I get the same thing. I sometimes get a passed result (although I don't think I've seen anything other than a black screen on the device).
What are the next steps?

Comment 11

[Aus Lacroix [:aus]]

I'm not sure what the question is here? The tests are run over a xvfb session (X window virtual frame buffer) so you won't see any output, that's normal.

We could change this if this is a requirement.

Otherwise, we'll continue to make things easier and more stable. We'd also like to do continuous integration on some select devices as well as CI MTBF.

This comment isn't meant to be a complete list of what's going to happen next. :)