Barrel. Using Erlang to build a Database

Barrel (https://barrel-db.org) is a modern document-oriented database in Erlang focusing on data locality (put/match the data next to you) and P2P with an effort to maintain a compatibility with the Apache CouchDB API.

In this talk, I will begin with an overview of Barrel, covering the problems that it is intended to solve and how its data model, transaction model, query model, and deployment model work together to solve those problems. I will then use Barrel to illustrate how to design and implement a production software in Erlang.

Transcript

1. BARREL
   DISTRIBUTED DATABASE
   FOR THE MODERN WORLD
   https://barrel-db.org
   Erlang Factory Lite Brussel 2016
   

   View Slide

2. VISION
   AND CONCEPT
   

   View Slide

3. local database
   mobile
   sensor
   "cloud" database
   local
   database
   mobile
   DATA IS MOBILE
   

   View Slide

4. share
   discover
   replicate
   PEER TO PEER (P2P)
   

   View Slide

5. ▸ Local first
   ▸ Put/Match the data next to you
   ▸ Query Locally
   ▸ Replicate a view of the data you need
   

   View Slide

6. WHAT
   IS BARREL
   

   View Slide

7. WHAT IS BARREL
   ▸ a document database
   ▸ revision tree & atomic document updates
   ▸ changes feed
   ▸ replication between any nodes in both way
   ▸ views (~ map)
   

   View Slide

8. AN EFFORT TO MAINTAIN
   THE COUCHDB API
   

   View Slide

9. ▸ view changes
   ▸ replication based on a view
   ▸ bulk get
   ▸ pluggable
   WITH MODERN FEATURES
   

   View Slide

10. ▸ DATA are not only blobs
    ▸ Replicated APPs
    ▸ Couchapps but extended and revisited
    REPLICATED APPS
    

    View Slide

11. BUILT
    IN ERLANG
    />
    <
    

    View Slide

12. ▸ append only file
    ▸ MVCC
    ▸ 2 indexes (btree): by sequence, by id
    ▸ 1 index for local documents without conflict handling
    ▸ A revision tree is stored in indexes pointed to the revision offset
    ▸ The revision is stored in the file separately
    HOW ARE STORED DOCUMENTS
    

    View Slide

13. REVISION TREE
    

    View Slide

14. ▸ Write is slow
    ▸ Read should not being blocked by writes
    ▸ No shared memory
    ▸ No atomic integer trick
    ▸ Only actors and message passing
    ▸ Operations on a doc are atomic
    CHALLENGES
    

    View Slide

15. READ/WRITE OPERATIONS
    DB STATE
    READER
    READER
    writer
    writer
    write process
    update
    share state
    

    View Slide

16. ▸ LRU to cache blocks

    https://github.com/barrel-db/erlang-lru
    ▸ 1 File process, Operations are limited
    ▸ DB users are linked to the database process
    ▸ Write buffer to reduce the latency
    ▸ Optional wal (barrel 014)
    ▸ NIF for the collation
    READ/WRITE OPERATIONS
    

    View Slide

17. INDEX OPERATIONS
    View Group
    READER
    READER
    change reader
    indexer
    update
    share state
    send /collect changes
    DB
    get changes
    

    View Slide

18. ▸ Credit Flow Based
    ▸ The View group keep the state
    ▸ View group is created on demande
    ▸ kept open until it has readers
    ▸ Indexer ask for updates
    ▸ Read functions (Map functions) are processed in //
    INDEX OPERATIONS
    

    View Slide

19. ▸ Create a new file to remove the fragmentation
    ▸ A race between copy and the addition of new data
    ▸ Require at least twice of the storage
    THE COMPACTION ISSUE
    

    View Slide

20. ▸ Added 2 features:
    ▸ MOVE: move doc(s) to another node or database
    ▸ User hooks functions (run in background) using hooks:

    https://github.com/barrel-db/hooks
    ▸ Partition on demand
    ▸ Decision depends on the application needs
    NEW FUNCTIONS
    

    View Slide

21. CHANGES HANDLER
    subscriber
    change dispatcher
    broadcast changes
    DB
    

    View Slide

22. ▸ changes load balancing
    ▸ consumer subscribe on patterns (delete, update, …)
    ▸ Create changes Load Balancer on demand
    ▸ Allows remote nodes to subscribe to a queue
    ▸ Based on primer (released on February 2016)
    CHANGES EVENTS
    

    View Slide

23. ▸ fetch the revisions and their attachments

    not present on the node
    ▸ continuous or not
    ▸ try to collect multiple docs at once
    ▸ use hackney:

    http://github.com/benoitc/hackney
    ▸ Use a Flow-based pattern instead of a classic pool
    REPLICATION
    

    View Slide

24. View Slide

25. HTTPS://BARREL-DB.ORG
    Barrel
    HTTP://ENKIM.EU
    Enki Multimedia
    

    View Slide