ElasticSearch Role

[Matheus Ashton]

Hi Guys,

First of all I want to congratulate you, provy is really awesome, in my company we use it in all our projects. One of our projects uses ElasticSearch for indexing and search.

I would like to contribute with the project creating an ElasticSearchRole, but I have some questions about it.

ElasticSearch doesn't have a package in debian nor in centos, to install it I have to download the TAR file and extract it. 

There is a specific way to do it (download the file, extract it to some dir etc) ? Some project standard or pattern that I must follow? I think it's a good thing creating one, because this is almost the same process that will be executed in "JettyRole" and "TomcatRole" issues that are open in provy's github. Solving one question, helps to solve them all.

In fact I can contribute with this two roles after I've made ElasticSearchRole.

Other question is that ElasticSearch depends on JRE, java must be installed on the same machine, how do you deal with this dependencies?

What do you think? 

Regards,

Matheus Ashton

[Fernando]

Hi Matheus,

On 6 February 2014 14:10, Matheus Ashton <matheu...@gmail.com> wrote:

Hi Guys,

First of all I want to congratulate you, provy is really awesome, in my company we use it in all our projects. One of our projects uses ElasticSearch for indexing and search.

I would like to contribute with the project creating an ElasticSearchRole, but I have some questions about it.

ElasticSearch doesn't have a package in debian nor in centos, to install it I have to download the TAR file and extract it. 

There is a specific way to do it (download the file, extract it to some dir etc) ? Some project standard or pattern that I must follow? I think it's a good thing creating one, because this is almost the same process that will be executed in "JettyRole" and "TomcatRole" issues that are open in provy's github. Solving one question, helps to solve them all.

In fact I can contribute with this two roles after I've made ElasticSearchRole.

I use to untar it on /tmp/ and delete it after all done (I also have used clean method if I have to access directory from different methods).

A example is nodejs role for debian, it just untar and runs make && make install.

I think the way we handle untar and install is not the most important, even if we have some standard way to do this (self.compile?), it would be hard, since each package needs something different.

Installing it the recommend way and cleaning up after are more important.

 

Other question is that ElasticSearch depends on JRE, java must be installed on the same machine, how do you deal with this dependencies?

with self.using(AptitudeRole) as role:

    role.ensure_package_installed('name-of-package')

Probably on provision method of the role.

 

[]s

What do you think? 

Regards,

Matheus Ashton

--
You received this message because you are subscribed to the Google Groups "provy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to provy+un...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--

Fernando Rocha, Software Developer
http://livingintheshell.com/ | ferna...@gmail.com

[Bernardo Heynemann]

Any news on this front?

Bernardo Heynemann
Developer @ globo.com