ELK stack Upgrade

I'm planning to upgrade my ELK stack for the latest versions:

  • logstash from 1.4.2 to 2.1.1
  • Elasticsearch from 1.4.4 para 2.1.0
  • Kibana from 4.1.2 para 4.3.0

Is there any documentation explaining the changes between releases for this products?

I've just found this ones:

Are there any others? Especially for ES versions between 1.4.4 and 2.0?

Additionally, is there any recomended method for upgrading ES? I think i'll have to do a full cluster restart because of the major release change... am i right?

Are there any others? Especially for ES versions between 1.4.4 and 2.0?

There are no release notes detailing the changes between versions X and Y. The release notes only detail changes between X and X+1. Therefore you should read the release notes for all releases between 1.4.4 and 2.0.

Additionally, is there any recomended method for upgrading ES? I think i'll have to do a full cluster restart because of the major release change... am i right?

Yes.

Also make sure you install and run the migration plugin in order to identify any potential issues before you start upgrading.

@magnusbaeck, where can i find the release notes for ES versions 1.5, 1.6 and 1.7? Changing the Elasticsearch referente on this page Release Notes | Elasticsearch Guide [2.1] | Elastic for this versions give me the following messages:

This page is not avaliable in the 1.7 version of the docs
This page is not avaliable in the 1.6 version of the docs
This page is not avaliable in the 1.5 version of the docs

@Christian_Dahlqvist, thanks a lot for the migration plugin tip!

Running the plugin on a dev cluster tells me that everything is just fine except for boolean fields on .kibana and .marvel index.

Is it safe to run the pluggin on a prod cluster?

Also, there is a link in the main page of the plugin for the breaking changes that may be what i was looking for!

where can i find the release notes for ES versions 1.5, 1.6 and 1.7?

You should be able to find them all from Past Releases of Elastic Stack Software | Elastic.

Thanks @magnusbaeck!