Mesos authentication

[pdread100]

Coded and tested Mesos authentication to resolve issue #35

[tnachen]

Please kill this new line

[pdread100]

You do not like the @SuppressWarnings("unchecked") ??

In general I like to suppress warnings that are not going to get fixed
since the warning in that case adds nothing and its annoying from a
completeness stand point.

But I will remove it.

Thanks

On Fri, May 1, 2015 at 5:26 PM, Timothy Chen notifications@github.com
wrote:

In src/storm/mesos/MesosNimbus.java
#37 (comment):

 } catch (Exception e) {
   throw new RuntimeException(e);
 }

}

• @SuppressWarnings("unchecked")
• protected void initialize(Map conf, String localDir) throws Exception {

Please kill this new line

—
Reply to this email directly or view it on GitHub
https://github.com/mesos/storm/pull/37/files#r29534976.

[pdread100]

Yes I see I should have added the issues in my answers I was responding to,
I'm use to threads ....but I will fix all issues.

Thanks.

On Tue, May 5, 2015 at 10:52 AM, Paul Read pdread101@gmail.com wrote:

You do not like the @SuppressWarnings("unchecked") ??

In general I like to suppress warnings that are not going to get fixed
since the warning in that case adds nothing and its annoying from a
completeness stand point.

But I will remove it.

Thanks

On Fri, May 1, 2015 at 5:26 PM, Timothy Chen notifications@github.com
wrote:

In src/storm/mesos/MesosNimbus.java
#37 (comment):

 } catch (Exception e) {
   throw new RuntimeException(e);
 }

}

• @SuppressWarnings("unchecked")
• protected void initialize(Map conf, String localDir) throws Exception {

Please kill this new line

—
Reply to this email directly or view it on GitHub
https://github.com/mesos/storm/pull/37/files#r29534976.

[erikdw]

@pdread100 : you misunderstood Tim. He meant "newline" rather than "new line". ;-) The annotation is fine.

[pdread100]

Ok. Thanks for the clarification.

[pdread100]

I have made all the changes requested by @tnachen. Is this sufficient? Thanks.

[erikdw]

hey @pdread100, irrespective of the outcome of @tnachen's response, can you please squash your commits into a single commit, and then force push? So that we get a clean single commit from this change. Thanks! If you need guidance on doing that let me know.

[erikdw]

@pdread100 : I just took a gander at your changes -- it seems that you missed @tnachen's request to not add all these getter/setter methods for variables which are only being used internally to the class. I agree with Tim, it's bloating the class unnecessarily.

[tnachen]

@erikdw I think I can squash them myself and remove the getter/setter since this has been sitting for a while. @pdread100 the changes looks good to me, but can you rebase on master? I cannot apply this PR on latest master.

[erikdw]

@tnachen : great, thanks!

[pdread100]

The last two days I have not had access to the net nor can I address this
until monday. Sorry for the not covering the issues but I thought the
setter/getter thing was not really an issue. I have been trying to squash
the commits but have not been too succesful.

Thanks

On Thu, May 14, 2015 at 7:30 PM, Erik Weathers notifications@github.com
wrote:

@tnachen https://github.com/tnachen : great, thanks!

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

[pdread100]

Ok. Rebased, squashed commits, removed the "protected" setters/getters (they were not private), should be good to go.

[brndnmtthws]

Whitespace.

[brndnmtthws]

Open a separate PR for the whitespace/formatting changes.

[maverick2202]

What username will the storm mesos framework use to run the storm topology ? Default, i think is root. Can we add it as a parameter too ?

[pdread100]

Mesos ACL will determine what the user tasks/topology is run as.

See Mesos Autorization
https://mesos.apache.org/documentation/latest/authorization/

On Tue, May 26, 2015 at 1:30 PM, maverick2202 notifications@github.com
wrote:

What username will the storm mesos framework use to run the storm topology
? Default, i think is root. Can we add it as a parameter too ?

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

[maverick2202]

Mesos ACL defines rules on how to run the task. The user account comes from the framework. In storm mesos framework, it is not set.

FrameworkInfo.Builder finfo = FrameworkInfo.newBuilder()
.........
.setUser("")

Can we also make it as a parameter ?

[pdread100]

I will double check but I believe you are incorrect. I believe mesos/myriad
is set up the exact same way an I think they do not set the user with the
explanation that mesos will handle adding the user.

On Wed, May 27, 2015 at 11:00 AM, maverick2202 notifications@github.com
wrote:

Mesos ACL defines rules on how to run the task. The user account comes
from the framework. In storm mesos framework, it is not set.

FrameworkInfo.Builder finfo = FrameworkInfo.newBuilder()
.........
.setUser("")

Can we also make it as a parameter ?

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

[erikdw]

The formatting is all totally out-of-whack after this change. The new code was written with 4-space indents, the old code was 2-space indents, and some of the functions run into each other (no newline in-between). I'll do a reformatting PR.

[pdread100]

I think we need a style checker, similar to what mesos/myriad uses. Such
that once code passes the style checker it should be good to go. I will
create a PR with an example one which can be tweaked to the satisfaction of
the group.

On Wed, May 27, 2015 at 2:50 PM, Erik Weathers notifications@github.com
wrote:

The formatting is all totally out-of-whack after this change. The new code
was written with 4-space indents, the old code was 2-space indents, and
some of the functions run into each other (no newline in-between). I'll do
a reformatting PR.

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

[pdread100]

I have created an issue to be worked in response to your request, issue
#43

On Wed, May 27, 2015 at 11:00 AM, maverick2202 notifications@github.com
wrote:

Mesos ACL defines rules on how to run the task. The user account comes
from the framework. In storm mesos framework, it is not set.

FrameworkInfo.Builder finfo = FrameworkInfo.newBuilder()
.........
.setUser("")

Can we also make it as a parameter ?

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

[erikdw]

Style checkers are better than nothing. Barely. They cannot / don't handle a lot of stuff, I personally hate them.

• Erik

On May 28, 2015, at 2:38 AM, pdread100 notifications@github.com wrote:

I think we need a style checker, similar to what mesos/myriad uses. Such
that once code passes the style checker it should be good to go. I will
create a PR with an example one which can be tweaked to the satisfaction of
the group.

On Wed, May 27, 2015 at 2:50 PM, Erik Weathers notifications@github.com
wrote:

The formatting is all totally out-of-whack after this change. The new code
was written with 4-space indents, the old code was 2-space indents, and
some of the functions run into each other (no newline in-between). I'll do
a reformatting PR.

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

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

[pdread100]

Yeah I'm not a fan either, however I have just started working with open
source projects and I'm running into the style buzz saw, some is legit and
a lot just absurd but it has to pass muster to get it into the repo.....
and it would be nice if there was some consistency from project to project.

I just ran two different style checkers and this project would require
massive "fixes" to get it to the level of other projects. And thats not
counting indents which is another sore spot. I simply don't understand how
anyone can justify 2 space indents...why even bother trying to make
something readable. Just make it straight line code and be done with it.
And the fear of a NL, holy cow bat man!

Anyway I will push a PR and folks can test it and if its not a good fit for
this project so be it.

On Thu, May 28, 2015 at 6:42 AM, Erik Weathers notifications@github.com
wrote:

Style checkers are better than nothing. Barely. They cannot / don't handle
a lot of stuff, I personally hate them.

• Erik

On May 28, 2015, at 2:38 AM, pdread100 notifications@github.com wrote:

I think we need a style checker, similar to what mesos/myriad uses. Such
that once code passes the style checker it should be good to go. I will
create a PR with an example one which can be tweaked to the satisfaction
of
the group.

On Wed, May 27, 2015 at 2:50 PM, Erik Weathers <notifications@github.com

wrote:

The formatting is all totally out-of-whack after this change. The new
code
was written with 4-space indents, the old code was 2-space indents, and
some of the functions run into each other (no newline in-between).
I'll do
a reformatting PR.

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

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

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

[erikdw]

Sounds great, thanks!

• Erik

On May 28, 2015, at 4:12 AM, pdread100 notifications@github.com wrote:

Yeah I'm not a fan either, however I have just started working with open
source projects and I'm running into the style buzz saw, some is legit and
a lot just absurd but it has to pass muster to get it into the repo.....
and it would be nice if there was some consistency from project to project.

I just ran two different style checkers and this project would require
massive "fixes" to get it to the level of other projects. And thats not
counting indents which is another sore spot. I simply don't understand how
anyone can justify 2 space indents...why even bother trying to make
something readable. Just make it straight line code and be done with it.
And the fear of a NL, holy cow bat man!

Anyway I will push a PR and folks can test it and if its not a good fit for
this project so be it.

On Thu, May 28, 2015 at 6:42 AM, Erik Weathers notifications@github.com
wrote:

Style checkers are better than nothing. Barely. They cannot / don't handle
a lot of stuff, I personally hate them.

• Erik

On May 28, 2015, at 2:38 AM, pdread100 notifications@github.com wrote:

I think we need a style checker, similar to what mesos/myriad uses. Such
that once code passes the style checker it should be good to go. I will
create a PR with an example one which can be tweaked to the satisfaction
of
the group.

On Wed, May 27, 2015 at 2:50 PM, Erik Weathers <notifications@github.com

wrote:

The formatting is all totally out-of-whack after this change. The new
code
was written with 4-space indents, the old code was 2-space indents, and
some of the functions run into each other (no newline in-between).
I'll do
a reformatting PR.

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

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

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

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