DEMO.md

MatchMiner V1.0 demo installation

This is a step by step guide to install and run a dockerized demo installation of MatchMiner using simulated data.

See demo_data directory

• Move genomic.bson, clinical.bson, trial.bson, and synonyms.txt to ./demo/matchminer/data/mongo

  Creating and setting up the images

1. Start the Docker Engine

   Check if the matchminer-demo-api:1.0.0 and matchminer-demo-ui:1.0.0 docker images are available. If not do step 2 - 5

   > docker images

   If they are present you should see:

   REPOSITORY                         TAG                 IMAGE ID            CREATED             SIZE
   matchminer-demo-ui                1.0.0               5c520c30478c        10 hours ago        1.04 GB
   matchminer-demo-api               1.0.0               938ea08d802c        10 hours ago        1.06 GB
   ...
   

   The UI image

   Navigate to matchminer-ui repository

2. > cd /matchminer-ui

3. > docker build -t matchminer-demo-ui:1.0.0 .

   The API image

   Navigate to matchminer-api repository

4. > cd /matchminer-api/

5. > docker build -t matchminer-demo-api:1.0.0 .

   NOTE: All mentioned paths are relative to the matchminer-api repository root.

   Building the remaining images (mongo, mongo-setup, kibana, mongo-connector, elasticsearch)

6. > cd ./demo/matchminer

7. > docker-compose up -d

   Check if the stack is running

8. > docker ps

   You should see something like the following:

   CONTAINER ID        IMAGE                                    COMMAND                  CREATED             STATUS              PORTS                                               NAMES
   ca04363b2487        matchminer-demo-ui:1.0.0                "/entrypoint.sh"         9 hours ago         Up 9 hours          0.0.0.0:8001->8001/tcp                              matchminer_ui_1
   cf7f2fb67265        matchminer-demo-kibana-sense:1.0.0      "/docker-entrypoin..."   9 hours ago         Up 9 hours          0.0.0.0:5601->5601/tcp                              matchminer_kibana_1
   326aa84fbf9b        matchminer-demo-mongo-connector:1.0.0   "/tmp/startup.sh"        9 hours ago         Up 9 hours                                                              matchminer_connector_1
   8a75218f7fca        matchminer-demo-api:1.0.0               "/entrypoint.sh dev"     9 hours ago         Up 9 hours          80/tcp, 443/tcp, 8443/tcp, 0.0.0.0:5555->5000/tcp   matchminer_api_1
   5b66b3b1c2c7        elasticsearch:2.4.0                     "/docker-entrypoin..."   9 hours ago         Up 9 hours          0.0.0.0:9200->9200/tcp, 9300/tcp                    matchminer_elastic_1
   896379773ce9        mongo:latest                            "docker-entrypoint..."   9 hours ago         Up 9 hours          27017/tcp                                           matchminer_mongo_1
   

   Check if the mongo replicateset (rs) has been initialized.

9. > docker exec -it matchminer_mongo_1 sh

10. # mongo

You should see the following prompt presented:

rs0:PRIMARY>

If this is the case then the replicateset has been initialized. If this has not been done correctly you would see rs0:OTHER> as a prompt. To initialize the replicateset run the following command:

 `> rs.initiate();`

Loading the data

11. Exit the mongo shell (> exit or ctrl-c)
12. > cd /storage

Loading the genomic data

13. > mongorestore --db matchminer genomic.bson

Loading the clinical data

14. > mongorestore --db matchminer clinical.bson

    The clinical trials will be added later. We will have to correctly add the mapping to Elasticsearch first.

Insert a demo user

15. > cat /storage/user.txt

    Copy the first json object into the clipboard and insert it into the mongo database

16. > mongo

17. # use matchminer;

18. db.user.insert(PASTE JSON OBJECT HERE)

    Find the _id of the user just inserted

19. db.user.find()

    Copy the _id of the user to the clipboard

20. Exit the mongo shell and exit the container (> exit (2x))

    Add the demo user _id to the config.json of the UI

21. > docker exec -it matchminer_ui_1 sh

    NOTE: You might have to install a text editor if its not present in the container

    Installing a text editor

22. > apt-get update

23. > apt-get install vim

24. > cd properties

25. > vim config.json

    Replace the user_id under the following path: dev.ENV.devUser.user_id

26. Exit the UI container

    Restart the ui container

27. > cd ./demo/matchminer

28. > docker-compose restart ui

    Getting the Elasticsearch mapping right

    Bring down the mongo-connector

29. > docker-compose stop connector

    In browser navigate to: localhost:5601

30. Set Elasticsearch URL in kibana to: http://elasticsearch:9200

    Drop the matchminer Elasticsearch index

31. In the large text field on the left enter: DELETE matchminer

32. Click the green play icon on the top right of the text box

    Copy the mapping JSON structure that came with this installation guide (./demo/data/mongo/elasticsearch_settings.txt)

33. In the same text box add POST matchminer { PASTE JSON MAPPING HERE }

34. Click the green play button to execute the command

    Launch the mongo container shell again

35. > docker exec -it matchminer_mongo_1 sh

36. # cd /storage

37. # mongorestore --db matchminer trial.json

38. Exit the mongo container (# exit)

    Start the mongo-connector

39. > docker-compose start connector

    All should be well! Check if the UI is running correctly in your browser:

http://localhost:8001