THIS IS WORK IN PROGRESS AND CAN CHANGE ANY TIME (wink)

Overview

At the moment we are developing a simple plugin for odysseus. 

To try the current state you could use the following docker file:

FROM node:14 as node-base
RUN echo "Installing required packages..." && \
	apt-get update && \
	apt-get install -yq --no-install-recommends python
RUN echo "Cloning Odysseus Grafana plugin source..." && \
	git clone https://git.swl.informatik.uni-oldenburg.de/scm/api_apps/grafana.git /var/lib/ody_grafana
WORKDIR /var/lib/ody_grafana/grafana-plugins/odysseus
RUN mkdir build
RUN echo "Building Odysseus Grafana plugin binaries..." && \
	yarn install && \
	yarn dev
	
FROM grafana/grafana:7.4.3 AS grafana
RUN echo "Copying Odysseus Grafana plugin binaries..."
COPY --from=node-base /var/lib/ody_grafana/grafana-plugins/odysseus /var/lib/grafana/plugins/ody-streaming-datasource
WORKDIR /usr/share/grafana
USER grafana
ENTRYPOINT [ "/run.sh" ]

Use this:

docker build . -t grafana
docker run -d -p 3000:3000 --name grafana grafana


Then you could use Odysseus in grafana as a datasource.

You will need to provide the server-adress: e.g. with ws://localhost:8888/

ATM there are some shortcommings that will be removed in future versions:

You will need to determine the websocket address manually by using a webbrowser (we recommend Firefox because of better JSON presentation) (http://localhost:8888/queries/<nr>) and find the section websockets with protocol JSON and use the entry with uri for creating panels in grafana:


The output of the connected query

  • should contain an attribute named time with a milliseconds timestamp (typically this should be a current timestamp for grafana)
  • all other attributes should be numeric attributes


A short video


The video can also be found here https://odysseus.informatik.uni-oldenburg.de/download/documentation/videos/grafana/odysseus_grafana_tutorial.mp4

In addition, here is a brief READ.ME.

We need a grafana plugin to be able to receive data from Odysseus.
Note that the current state of the plugin is experimental.

Wiki-Page: https://wiki.odysseus.informatik.uni-oldenburg.de/display/ODYSSEUS/Grafana

Requirements: docker

1. build a docker image based on the dockerfile in this folder
docker build . -t grafana

2. build and run a docker container based on the created image
docker run -d -p 3000:3000 --name grafana grafana

3. open grafana in a browser
localhost:3000
username: admin
password: admin

4. Add an Odysseus datasource
Requirements: Running Odysseus Server (or monolithic of course)
Configuration -> Data Sources
Select Others -> odysseus
Server URI: ws://localhost:8888/ (REST-port of a (here local) odysseus server)
Leave API Key empty

5. Create Dashboard
Create -> Dashboard

6. Create Panel
Requirements:
* To visualize single datastream attributes, there must be an operator that provides only the current timestamp named "time" and that attribute
* To easily find the operator, give him a proper name
* Query(s) should be installed
Tipps:
a) At the end of the query, create for each attribute you want to visualize a projection on "time" and the attribute
b) Grafana can display historical values but to show the results in real-time, the timestamps of the results should be set to the current time (necessary if you, e.g., read from a CSV file with old timestamps)
6.1. select your Odysseus datasource as datasource
6.2. you need a websocket that provides the data
6.3. open in a browser localhost:8888/queries to get information about all installed queries (in chrome it is not formatted; copy the content in, e.g., https://jsonformatter.curiousconcept.com/)
6.4. find the websocket in the json
go to the query id of your query -> rootOperators -> go to the operator providing the data to visualize -> ports -> websockets -> copy the JSON uri
6.5. Insert the copied URI in grafana into the websocket field of the panel query
6.6. Click on apply and after that on save dashboard

Development 

See https://grafana.com/docs/grafana/latest/developers/plugins/

and https://grafana.com/docs/grafana/latest/developers/plugins/build-a-streaming-data-source-plugin/ for further information how to extend the current implementation.

Quick start.

  • Install e.g. MS Code for development
  • Install anything necessary for grafana plugin development 
  • create a folder named /grafana-plugins
  • inside this folder clone the plugin
  • switch into plugin folder and run:

    yarn install
    yarn dev
  • now start grafana e.g. with
    docker run -d -e GF_DEFAULT_APP_MODE=development -p 3000:3000 -v "$(pwd)"/grafana-plugins:/var/lib/grafana/plugins --name=grafana grafana/grafana:latest
  • If you change anything inside the plugin, grafana needs to be restart. You could create a script:

    docker stop grafana
    docker rm grafana
    docker run -d -e GF_DEFAULT_APP_MODE=development -p 3333:3000 -v "$(pwd)"/grafana-plugins:/var/lib/grafana/plugins --name=grafana grafana/grafana:latest


  • No labels