Changes in nodes api in ES 1.0.0.RC1

Question

Changes in nodes api in ES 1.0.0.RC1

michaelsalmon opened this issue 11 years ago · 18 comments

The node info is fetched using /_nodes in RC1 and not /_cluster/nodes as in previous versions. This appears to only be used in NodeInfo.js line 25

Answer 1 · 2014-01-19T08:31:32.000Z

Yes, changing that line fixes the issue with no node charts on ES RC1.

Answer 2 · 2014-01-20T09:07:44.000Z

Thanks guys, I will have a look at this during this week. I will try to release updated version by the end of this week.

Answer 3 · 2014-02-04T17:02:29.000Z

Hi,
Just want to say that although I didn't notice your comment in the README about installing on ES 1.0 until later, based on your screenshots I can't wait for your improvements! Am curious though, it seems you didn't create a "ES 1.0 branch"? I assume then that as you merge patches to support 1.0 then the current Master which supports <1.0 will be lost unless someone wants to go to the trouble of undoing the patches?

Answer 4 · 2014-02-04T17:35:07.000Z

Hi,

there are coming bunch of changes (and I am pretty exited about them). But no worries, no history or past changes will be lost. I will fork existing master into new branch (something like 0.90.x) before I commit 1.0.x code into master so we can still bugfix for previous versions without any issues.

I am working on new(*) code pretty hard these days and I can't wait to see it released. It is not in public branch though because I want to be sure that the code is settled correctly. You can be sure I am pretty serious about performance and robustness of the new code. I am fully focused on it now, so if anybody is up to patching current code, feel free to send PR. There are still some hard coding nights in front of me before this thing can go public... (thanx for patience).

(*) when I say new I really mean it! 😄

Answer 5 · 2014-02-10T11:20:25.000Z

Hi @lukas-vlcek

Thanks for maintaining Bigdesk. It's a great piece of software.

Is there something we can use while you work on the changes in ES 1.0 ?

I'd like to have some kind of view about how the cluster behaves after an upgrade (I use 1.0.0.RC2 right now).

Thanks

Answer 6 · 2014-02-10T12:11:11.000Z

I think there are options, if I am not mistaken Elasticsearch-HQ supports 1.0.x already (//cc @royrusso ?) and of course Marvel. But I am not much familiar with any of these so can not give you exact comparison. There might be also other options but I am not sure how much they already support 1.0.x (see http://www.elasticsearch.org/guide/en/elasticsearch/client/community/current/health.html). Also kopf might be an option too (? - I am listing kopf here because I did not find a reference on es.org site).

One thing I can say about the new Bigdesk right now is that in this release it won't compete with tools that can store ES stats for a long period of time. This means, if you really need long time persistency of data for later investigation, don't hold your breath and switch to relevant options now.

HTH

Answer 7 · 2014-02-10T12:23:30.000Z

However, I would like to point out that I am really amazed about the feedback I get about Bigdesk. It really started as a small pet project couple years ago and I wouldn't have imagined this going so far. I take it much more seriously now.

Answer 8 · 2014-02-10T12:24:40.000Z

I changed _cluster/nodes to _nodes on line 25 of NodeInfo.js and I haven’t noticed any problems so far.

On 10 Feb 2014, at 12:20:28, Jérémy Lecour notifications@github.com wrote:

Hi @lukas-vlcek

Thanks for maintaining Bigdesk. It's a great piece of software.

Is there something we can use while you work on the changes in ES 1.0 ?

I'd like to have some kind of view about how the cluster behaves after an upgrade (I use 1.0.0.RC2 right now).

Thanks

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

/Michael Salmon
Efficiency IT Expert
Inovia AB

Answer 9 · 2014-02-10T12:41:03.000Z

@michaelsalmon that sounds like pretty simple fix. Would you mind doing a small PR?

Answer 10 · 2014-02-10T12:45:31.000Z

I’m pretty new to github, by PR do you mean fork, edit, pull request?

On 10 Feb 2014, at 13:41:06, Lukáš Vlček notifications@github.com wrote:

@michaelsalmon that sounds like pretty simple fix. Would you mind doing a small PR?

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

/Michael Salmon
Efficiency IT Expert
Inovia AB

Answer 11 · 2014-02-10T12:51:44.000Z

@lukas-vlcek I'm with you about Bigdesk not competing with Marvel (or else), but as you say, it remains pertinent for quick checks (or during a few minutes/hours). It is also much easier to use.

Also, @michaelsalmon's fix seems OK.

I've just made the PR

Answer 12 · 2014-02-10T13:15:39.000Z

You guys rock! Sounds like I have a new plan for today evening :-)

Answer 13 · 2014-02-10T18:21:46.000Z

At the moment,
I am running on 1.0 RC1
elasticsearch-head
elasticsearch-hq (Almost everything seems to be working. Just a few items
read null)
Marvel

Tony

On Mon, Feb 10, 2014 at 4:11 AM, Lukáš Vlček notifications@github.comwrote:

I think there are options, if I am not mistaken Elasticsearch-HQhttps://github.com/royrusso/elasticsearch-HQsupports 1.0.x already (//cc
@royrusso https://github.com/royrusso ?) and of course Marvelhttp://www.elasticsearch.org/overview/marvel/.
But I am not much familiar with any of these so can not give you exact
comparison. There might be also other options but I am not sure how much
they already support 1.0.x (see
http://www.elasticsearch.org/guide/en/elasticsearch/client/community/current/health.html).
Also kopf https://github.com/lmenezes/elasticsearch-kopf might be an
option too (? - I am listing kopf here because I did not find a reference
on es.org site).

One thing I can say about the new Bigdesk right now is that in this
release it won't compete with tools that can store ES stats for a long
period of time. This means, if you really need long time persistency of
data for later investigation, don't hold your breath and switch to relevant
options now.

HTH

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/46#issuecomment-34624848
.

Answer 14 · 2014-02-10T20:35:25.000Z

i've updated my cluster to use OpenJDK 1.7.0_25, on Debian 7, with Elasticsearch 1.0.0.RC2 for both nodes.

I get many missing metrics, but overall its better than nothing.

Answer 15 · 2014-02-21T19:40:30.000Z

Bigdesk 2.4.0 has been just released. Check announcement details https://groups.google.com/d/msg/elasticsearch/Q1mLZI6iBDU/smc_oatwkDgJ
Closing...

Answer 16 · 2014-02-22T09:18:27.000Z

How to d./l?

I had 1.x already downloaded (git clone).
I did a git pull, received some updated file but when I copied to a
webserver and pointed to a node ES 1.0 GA, All I get is the intial
"connect" page and the experimental cluster diagram. Everything else
refuses to work.

I don't see how to check current deployed version and it's not displayed.
Also, when I inspect the Git repo, it looks like there are only the Main
branch plus some branches which clearly are 1.x bigdesk when I should be
looking for a 2.x.... So, I am assuming that "main" is the 2.x bigdesk
branch.

Tony

On Fri, Feb 21, 2014 at 11:40 AM, Lukáš Vlček notifications@github.comwrote:

Closed #46 #46.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/46
.

Answer 17 · 2014-02-22T13:42:30.000Z

Hi,

what you should be interested in are git tags not branches. Branches are used for development, experiments and sometimes forks of old bigdesk versions. Below you can find some examples how to get Bigdesk. BTW with future Bigdesk releases (promised version 3.) whole git repo will be structured a bit differently, hopefully it will be more clear what is the development and what are final releases.

If you want to install and use Bigdesk 2.4.0 (or any older version) then there are couple of ways how to do it. Some of them includes:

1) Use it as an Elasticsearch plugin

Install Elasticsearch 1.0.0 then navigate to <ES_HOME> and execute the following command:

$ ./bin/plugin -install lukas-vlcek/bigdesk/2.4.0

In my case it does the following:

-> Installing lukas-vlcek/bigdesk/2.4.0...
Trying http://download.elasticsearch.org/lukas-vlcek/bigdesk/bigdesk-2.4.0.zip...
Trying http://search.maven.org/remotecontent?filepath=lukas-vlcek/bigdesk/2.4.0/bigdesk-2.4.0.zip...
Trying https://oss.sonatype.org/service/local/repositories/releases/content/lukas-vlcek/bigdesk/2.4.0/bigdesk-2.4.0.zip...
Trying https://github.com/lukas-vlcek/bigdesk/archive/v2.4.0.zip...
Downloading ........................................................................................................................................................................................................................................................DONE
Installed lukas-vlcek/bigdesk/2.4.0 into /Users/lukas/projects/elasticsearch-1.0.0/plugins/bigdesk
Identified as a _site plugin, moving to _site structure ...

Then you can open web browser and navigate to http://localhost:9200/_plugin/bigdesk it should open installed Bigdesk and automatically auto-connect to local node and redirect to http://localhost:9200/_plugin/bigdesk/#nodes.

2) Clone Bigdesk repo into localhost and checkout specific tag

Clone Bigdesk repo (the master branch) and checkout specific release tag.

$ git clone https://github.com/lukas-vlcek/bigdesk.git
Cloning into 'bigdesk'...
remote: Counting objects: 4947, done.
remote: Compressing objects: 100% (2726/2726), done.
remote: Total 4947 (delta 1831), reused 4883 (delta 1792)
Receiving objects: 100% (4947/4947), 17.78 MiB | 1.08 MiB/s, done.
Resolving deltas: 100% (1831/1831), done.

$ cd bigdesk/
$ git tag
[... some tags left out for brevity ....]
v2.2.2
v2.2.3
v2.4.0

$ git checkout v2.4.0
Note: checking out 'v2.4.0'.
[... optional git messages ...]
HEAD is now at 4a23042... Release v2.4.0 - Extract the cluster status CSS and make it more readable

Now you should be able to open index.html and point it to Elasticsearch node.

3) You can download archives from GitHub

See http://bigdesk.org/#download_locally
BTW these archives are provided by GitHub and are based on above discussed git tags.

4) You can use Bigdesk online

See http://bigdesk.org/v/

Answer 18 · 2014-02-22T16:20:55.000Z

Thx!

Tony
On Feb 22, 2014 5:42 AM, "Lukáš Vlček" notifications@github.com wrote:

Hi,

what you should be interested in are git tags not branches. Branches are
used for development, experiments and sometimes forks of old bigdesk
versions. Below you can find some examples how to get Bigdesk. BTW with
future Bigdesk releases (promised version 3.) whole git repo will be
structured a bit differently, hopefully it will be more clear what is the
development and what are final releases.

If you want to install and use Bigdesk 2.4.0 (or any older version) then
there are couple of ways how to do it. Some of them includes:

Use it as an Elasticsearch plugin

Install Elasticsearch 1.0.0 then navigate to <ES_HOME> and execute the
following command:

$ ./bin/plugin -install lukas-vlcek/bigdesk/2.4.0

In my case it does the following:

-> Installing lukas-vlcek/bigdesk/2.4.0...
Trying http://download.elasticsearch.org/lukas-vlcek/bigdesk/bigdesk-2.4.0.zip...
Trying http://search.maven.org/remotecontent?filepath=lukas-vlcek/bigdesk/2.4.0/bigdesk-2.4.0.zip...
Trying https://oss.sonatype.org/service/local/repositories/releases/content/lukas-vlcek/bigdesk/2.4.0/bigdesk-2.4.0.zip...
Trying https://github.com/lukas-vlcek/bigdesk/archive/v2.4.0.zip...
Downloading ........................................................................................................................................................................................................................................................DONE
Installed lukas-vlcek/bigdesk/2.4.0 into /Users/lukas/projects/elasticsearch-1.0.0/plugins/bigdesk
Identified as a _site plugin, moving to _site structure ...

Then you can open web browser and navigate to
http://localhost:9200/_plugin/bigdesk it should open installed Bigdesk
and automatically auto-connect to local node and redirect to
http://localhost:9200/_plugin/bigdesk/#nodes.
2) Clone Bigdesk repo into localhost and checkout specific tag

Clone Bigdesk repo (the master branch) and checkout specific release tag.

$ git clone https://github.com/lukas-vlcek/bigdesk.git
Cloning into 'bigdesk'...
remote: Counting objects: 4947, done.
remote: Compressing objects: 100% (2726/2726), done.
remote: Total 4947 (delta 1831), reused 4883 (delta 1792)
Receiving objects: 100% (4947/4947), 17.78 MiB | 1.08 MiB/s, done.
Resolving deltas: 100% (1831/1831), done.

$ cd bigdesk/
$ git tag
[... some tags left out for brevity ....]
v2.2.2
v2.2.3
v2.4.0

$ git checkout v2.4.0
Note: checking out 'v2.4.0'.
[... optional git messages ...]
HEAD is now at 4a23042... Release v2.4.0 - Extract the cluster status CSS and make it more readable

Now you should be able to open index.html and point it to Elasticsearch
node.
3) You can download archives from GitHub

See http://bigdesk.org/#download_locally
BTW these archives are provided by GitHub and are based on above discussed
git tags.
4) You can use Bigdesk online

See http://bigdesk.org/v/

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/46#issuecomment-35802813
.