Bump to Ninja 3.3.1 - important drop in upgrade.

[raphaelbauer]

Also removed no longer necessary exclusions for guava. Painless drop in replacement.

[mcocciaTE]

I pulled your changes and somehow it broke the framework for me. All of the tests in both frameworks are printing 500 Internal Server Errors. There might be an issue with how things are implemented in 3.1.4 vs how they were in 2.5.1. I also tried updating everything to 3.2.0 but that did not work either.

[raphaelbauer]

Hey @mcocciaTE ,

I fixed the problems. It should work now without any problems (mvn clean install). The problem was caused by the new case of "Hello, World!" in json/plaintext responses,

Thanks for your efforts!

Best,

Raphael

[aschneider-techempower]

Unfortunately, all of the tests fail for ninja and ninja-standalone.

Here is the error log for ninja-standalone (ninja didn't have any errors that were in the log)

[raphaelbauer]

Do the ninja-standalone tests run? (I bet they do)....

Do you have any detailed error message from resin what is going on?

I mean - they already ran in the previous benchmarks - so it can only be a
tiny config issue...

Thanks for your support!

Raphael

On Fri, Jul 18, 2014 at 9:47 PM, Alexander Schneider <
notifications@github.com> wrote:

Unfortunately, all the tests fail
https://gist.github.com/aschneider-techempower/702637232d52084ccad7.

—
Reply to this email directly or view it on GitHub
#800 (comment)
.

inc: http://ars-machina.raphaelbauer.com
tech: http://ars-codia.raphaelbauer.com
web: http://raphaelbauer.com

[aschneider-techempower]

The ninja and ninja-standalone tests as they are right now run and pass all validation steps.

Here is the Resin log. Some of the ones are from Ninja running pre-PR and some are running on the PR.

[raphaelbauer]

On Fri, Jul 18, 2014 at 10:08 PM, Alexander Schneider <
notifications@github.com> wrote:

The ninja and ninja-standalone tests as they are right now run and pass
all validation steps.

Aeh... Just as clarification "ninja" and "ninja-standalone" tests run...
ok. And which tests do no run (path in the git repo)? (Because there should
only be two ninja projects at the moment)...

Maybe a stupid question - but maybe we debug something that isn't really a
problem...

Many thanks!

Raphael

[aschneider-techempower]

Prior to the pull request, the tests run fine. After the pull requests, the tests fail.

[raphaelbauer]

The logfles are from an old Ninja version run (Ninja 2.5.1) and also
include logs from "wicket" (another java framework).

The current version of the code is this one
https://github.com/raphaelbauer/FrameworkBenchmarks/blob/master/ninja-resin/pom.xml

And Ninja is version 3.2.0.

Ninja prints out asciiart with the version:

_______ .___ _______ . _____
\ \ | |\ \ | | / _
/ | | |/ | \ | |/ /\
/ | \ / | /| / | \ http://www.ninjaframework.org
__| /****|__ /________| / @ninjaframework
web/framework / / 2.5.1

If you see "2.5.1" it's NOT the new pull request. It should say "3.2.0".

I hope that helps somehow...

I hope it's not too frustrating for you. Many thanks for your effort
and let me know if I can help...

Thanks!

Raphael

On Fri, Jul 18, 2014 at 10:19 PM, Alexander Schneider
notifications@github.com wrote:

Prior to the pull request, the tests run fine. After the pull requests, the
tests fail.

—
Reply to this email directly or view it on GitHub.

inc: http://ars-machina.raphaelbauer.com
tech: http://ars-codia.raphaelbauer.com
web: http://raphaelbauer.com

[aschneider-techempower]

The reason it includes wicket is because I was also merging a wicket test.

Here are the logs JUST from your version of ninja. If there's a version disparity, it needs to be resolved in the setup.py file.

[raphaelbauer]

Okay... thanks that helps a lot... does the "other" ninja testproject run
(ninja-standalone)?

Thanks,

Raphael

On Fri, Jul 18, 2014 at 10:45 PM, Alexander Schneider <
notifications@github.com> wrote:

The reason it includes wicket is because I was also merging a wicket test.

Here https://gist.github.com/aschneider-techempower/9230681d863c67dfb653
are the logs JUST from your version of ninja. If there's a version
disparity, it needs to be resolved in the setup.py file.

—
Reply to this email directly or view it on GitHub
#800 (comment)
.

inc: http://ars-machina.raphaelbauer.com
tech: http://ars-codia.raphaelbauer.com
web: http://raphaelbauer.com

[aschneider-techempower]

Neither ninjas from your pull request run.

[raphaelbauer]

Could you please provide me with the error logs of ninja-standalone?

It can only be something tiny related to the setup on the benchmarking
machines...

Thanks!

Raphael

On Fri, Jul 18, 2014 at 11:09 PM, Alexander Schneider <
notifications@github.com> wrote:

Neither ninjas from your pull request run.

—
Reply to this email directly or view it on GitHub
#800 (comment)
.

inc: http://ars-machina.raphaelbauer.com
tech: http://ars-codia.raphaelbauer.com
web: http://raphaelbauer.com

[aschneider-techempower]

The error logs have already been provided. They are here

[raphaelbauer]

But these are not the complete logs afaics - or am I wrong?

Raphael

On Fri, Jul 18, 2014 at 11:23 PM, Alexander Schneider <
notifications@github.com> wrote:

The error logs have already been provided. They are here
https://gist.github.com/aschneider-techempower/bd96f5b107fa0872730e

—
Reply to this email directly or view it on GitHub
#800 (comment)
.

inc: http://ars-machina.raphaelbauer.com
tech: http://ars-codia.raphaelbauer.com
web: http://raphaelbauer.com

[aschneider-techempower]

These are all the logs from stderr. Are there other logs that I could provide you?

[raphaelbauer]

Okay... strange strange... I am trying to reproduce the problem here...
Must be something with the setup step...
Give me some minutes...
Thanks,
ra

On Fri, Jul 18, 2014 at 11:26 PM, Alexander Schneider <
notifications@github.com> wrote:

These are all the logs from stderr. Are there other logs that I could
provide you?

—
Reply to this email directly or view it on GitHub
#800 (comment)
.

inc: http://ars-machina.raphaelbauer.com
tech: http://ars-codia.raphaelbauer.com
web: http://raphaelbauer.com

[martin-g]

You are welcome!

BTW I am not TE employee! Apply my changes to your repo and branch if you
didn't already so they appear in the pull request..
On Jul 23, 2014 11:40 PM, "Raphael A. Bauer" notifications@github.com
wrote:

ok...
so

1. ninja-standalone works
2. ninja-resin now works, too
3. debug level is back to debug

Did I get this right?

—
Reply to this email directly or view it on GitHub
#800 (comment)
.

[raphaelbauer]

Hey @martin-g - that's awesome! I added you to README.md as contributor to say thanks!

Really good work.

@mcocciaTE and @aschneider-techempower ninja-standalone and ninja-resin should work by now. Would be great if you could confirm...

Thanks for all your support. Really stunned by your great support!

[raphaelbauer]

@mcocciaTE and @aschneider-techempower bumping that back on top - would be great if you could verify and close this PR! Thanks for your support!

[aschneider-techempower]

You're still failing one of the tests.

-----------------------------------------------------
  VERIFYING UPDATE (/ninja/update?queries=)
-----------------------------------------------------
HTTP/1.1 500 Internal Server Error
Server: Resin/4.0.36
Expires: Thu, 01 Jan 1970 00:00:00 GMT
Cache-Control: no-cache, no-store, max-age=0, must-revalidate
Content-Type: text/html; charset=utf-8
Transfer-Encoding: chunked
Date: Tue, 29 Jul 2014 15:13:54 GMT

<html>
<head>
<title>Oops. Thats an internal server error and all we know.</title>
</head>
<body>
    <h1>Oops. Thats an internal server error and all we know.</h1>
</body>
</html>






VALIDATING UPDATE ... 
FAIL Got exception when trying to validate the update test: None

JVM Resin Log

[raphaelbauer]

I made a change i the persistence layer - could you please try that again?
If that still does not work I'd propose to deactivate ninja-resin and only run the benchmarks with ninja-standalone.

Thanks for your efforts!

[raphaelbauer]

@aschneider-techempower and @mcocciaTE - I bumped ninja to the latest 3.3.1.

Just wondering if you could merge the PR for the next round. ninja-standalone should run by now. And if ninja-resin does not run I'd suggest to disable it :)

Thanks for your efforts!

[hamiltont]

apologies, but could you rebase again? I had to apply a hotfix to our travis setup

[raphaelbauer]

I synced the fork... Let me know if there is anything else to do :) Thanks for your efforts!

[hamiltont]

@raphaelbauer could you please merge in my ninja-800 branch? From my initial tests there appear to be a number of http 500s occurring.

Also, do you expect a number of additional frameworks to depend on resin? If not (e.g. ninja-resin will basically be the only directory that needs to install resin), then could you move the code from resin.sh into ninja-resin/install.sh below the fw_depends? This will mean that you can change/update the resin installation anytime without triggering Travis to verify every framework we have ;-)

[raphaelbauer]

Hey @hamiltont

• I deleted the ninja-resin and pushed my PR. It's just too cumbersome for me to maintain - and I am not a resin Pro. Ninja still is included and running happily as ninja-standalone.

Could you verify that that's working?

Thanks!

[hamiltont]

I very much understand. Perhaps we can open an issue or broadcast on the
google groups that we could use a maintainer for the resin setup if someone
wants to see it included.

Re standalone, it passed verify fine last time so I expect no problems, but
I'll wait to see if travis says we are still good before merging.

[hamiltont]

Oh ps - the .travis.yml file should have the line for ninja-resin removed at soem point

[hamiltont]

Travis checks out, this LGTM. Thanks @raphaelbauer !

[raphaelbauer]

hey @hamiltont - I removed ninja-resin from travis. Regarding support from the community - that's a very good idea. Actually ninja-standalone and ninja-resin were already created by "the community". I'll ask on the mailing lists... Thanks for you efforts... Well appreciated!