An Alternative Elixir Driver for MongoDB

Features

Supports MongoDB versions 3.2, 3.4, 3.6, 4.0, 4.2, 4.4
Connection pooling (through DBConnection 2.x)
Streaming cursors
Performant ObjectID generation
Aggregation pipeline
Replica sets
Support for SCRAM-SHA-256 (MongoDB 4.x)
Support for GridFS (See)
Support for change streams api (See)
Support for bulk writes (See)
support for driver sessions (See)
support for driver transactions (See)
support for command monitoring (See)
support for retryable reads (See)
support for retryable writes (See)

Data Representation

BSON                Elixir
----------          ------
double              0.0
string              "Elixir"
document            [{"key", "value"}] | %{"key" => "value"} (1)
binary              %BSON.Binary{binary: <<42, 43>>, subtype: :generic}
object id           %BSON.ObjectId{value: <<...>>}
boolean             true | false
UTC datetime        %DateTime{}
null                nil
regex               %BSON.Regex{pattern: "..."}
JavaScript          %BSON.JavaScript{code: "..."}
integer             42
symbol              "foo" (2)
min key             :BSON_min
max key             :BSON_max
decimal128          Decimal{}

Since BSON documents are ordered Elixir maps cannot be used to fully represent them. This driver chose to accept both maps and lists of key-value pairs when encoding but will only decode documents to lists. This has the side-effect that it's impossible to discern empty arrays from empty documents. Additionally the driver will accept both atoms and strings for document keys but will only decode to strings.

BSON symbols can only be decoded.

Usage

Installation

Add mongodb_driver to your mix.exs deps.

defp deps do
  [{:mongodb_driver, "~> 0.7"}]
end

Then run mix deps.get to fetch dependencies.

Simple Connection to MongoDB

# Starts an unpooled connection
{:ok, conn} = Mongo.start_link(url: "mongodb://localhost:27017/db-name")

# Gets an enumerable cursor for the results
cursor = Mongo.find(conn, "test-collection", %{})

cursor
|> Enum.to_list()
|> IO.inspect

To specify a username and password, use the :username, :password, and :auth_source options.

# Starts an unpooled connection
{:ok, conn} = 
    Mongo.start_link(url: "mongodb://localhost:27017/db-name", 
                     username: "test_user", 
                     password: "hunter2",
                     auth_source: "admin_test")

# Gets an enumerable cursor for the results
cursor = Mongo.find(conn, "test-collection", %{})

cursor
|> Enum.to_list()
|> IO.inspect

For secure requests, you may need to add some more options; see the "AWS, TLS and Erlang SSL ciphers" section below.

Failing operations return a {:error, error} tuple where error is a Mongo.Error object:

{:error,
 %Mongo.Error{
   code: 13435,
   error_labels: [],
   host: nil,
   message: "not master and slaveOk=false",
   resumable: true,
   retryable_reads: true,
   retryable_writes: true
 }}

Connection Pooling

The driver supports pooling by DBConnection (2.x). By default mongodb_driver will start a single connection, but it also supports pooling with the :pool_size option. For 3 connections add the pool_size: 3 option to Mongo.start_link and to all function calls in Mongo using the pool:

# Starts an pooled connection
{:ok, conn} = Mongo.start_link(url: "mongodb://localhost:27017/db-name", pool_size: 3)

# Gets an enumerable cursor for the results
cursor = Mongo.find(conn, "test-collection", %{})

cursor
|> Enum.to_list()
|> IO.inspect

If you're using pooling it is recommend to add it to your application supervisor:

def start(_type, _args) do
  import Supervisor.Spec

  children = [
    worker(Mongo, [[name: :mongo, database: "test", pool_size: 3]])
  ]

  opts = [strategy: :one_for_one, name: MyApp.Supervisor]
  Supervisor.start_link(children, opts)
end

Due to the mongodb specification, an additional connection is always set up for the monitor process.

Replica Sets

By default, the driver will discover the deployment's topology and will connect to the replica set automatically, using either the seed list syntax or the URI syntax. Assuming the deployment has nodes at hostname1.net:27017, hostname2.net:27017 and hostname3.net:27017, either of the following invocations will discover the entire deployment:

{:ok, pid} = Mongo.start_link(database: "test", seeds: ["hostname1.net:27017"])

{:ok, pid} = Mongo.start_link(url: "mongodb://hostname1.net:27017/test")

To ensure that the connection succeeds even when some of the nodes are not available, it is recommended to list all nodes in both the seed list and the URI, as follows:

{:ok, pid} = Mongo.start_link(database: "test", seeds: ["hostname1.net:27017", "hostname2.net:27017", "hostname3.net:27017"])

{:ok, pid} = Mongo.start_link(url: "mongodb://hostname1.net:27017,hostname2.net:27017,hostname3.net:27017/test")

Using an SRV URI also discovers all nodes of the deployment automatically.

Auth Mechanisms

For versions of Mongo 3.0 and greater, the auth mechanism defaults to SCRAM. If you'd like to use MONGODB-X509 authentication, you can specify that as a start_link option.

{:ok, pid} = Mongo.start_link(database: "test", auth_mechanism: :x509)

AWS, TLS and Erlang SSL Ciphers

Some MongoDB cloud providers (notably AWS) require a particular TLS cipher that isn't enabled by default in the Erlang SSL module. In order to connect to these services, you'll want to add this cipher to your ssl_opts:

{:ok, pid} = Mongo.start_link(database: "test", 
      ssl_opts: [
        ciphers: ['AES256-GCM-SHA384'],
        cacertfile: "...",
        certfile: "...")
      ]
)

Find

Using $and

Mongo.find(:mongo, "users", %{"$and" => [%{email: "[email protected]"}, %{first_name: "first_name"}]})

Using $or

Mongo.find(:mongo, "users", %{"$or" => [%{email: "[email protected]"}, %{first_name: "first_name"}]})

Using $in

Mongo.find(:mongo, "users", %{email: %{"$in" => ["[email protected]", "[email protected]"]}})

Change Streams

Change streams are available in replica set and sharded cluster deployments and tell you about changes to documents in collections. They work like endless cursors.

The special thing about change streams is that they are resumable: in case of a resumable error, no exception is propagated to the application, but instead the cursor is re-scheduled at the last successful location.

The following example will never stop, thus it is a good idea to use a process for reading from change streams:

seeds = ["hostname1.net:27017", "hostname2.net:27017", "hostname3.net:27017"]
{:ok, top} = Mongo.start_link(database: "my-db", seeds: seeds, appname: "getting rich")
cursor =  Mongo.watch_collection(top, "accounts", [], fn doc -> IO.puts "New Token #{inspect doc}" end, max_time: 2_000 )  
cursor |> Enum.each(fn doc -> IO.puts inspect doc end)

An example with a spawned process that sends messages to the monitor process:

def for_ever(top, monitor) do
    cursor = Mongo.watch_collection(top, "users", [], fn doc -> send(monitor, {:token, doc}) end)
    cursor |> Enum.each(fn doc -> send(monitor, {:change, doc}) end)
end

spawn(fn -> for_ever(top, self()) end)

For more information see:

Mongo.watch_collection

Inserts

To insert a single document:

Mongo.insert_one(top, "users", %{first_name: "John", last_name: "Smith"})

To insert a list of documents:

Mongo.insert_many(top, "users", [
  %{first_name: "John", last_name: "Smith"},
  %{first_name: "Jane", last_name: "Doe"}
])

Indexes

To create indexes you can call the function Mongo.createIndexed/4:

indexes =  [[key: [files_id: 1, n: 1], name: "files_n_index", unique: true]]
Mongo.create_indexes(topology_pid, "my_collection", indexes, opts)

You specify the indexes parameter as a keyword list with all options described in the documentation of the createIndex command.

For more information see:

Bulk Writes

The motivation for bulk writes lies in the possibility of optimization, the same operations to group. Here, a distinction is made between disordered and ordered bulk writes. In disordered, inserts, updates, and deletes are grouped as individual commands sent to the database. There is no influence on the order of the execution. A good use case is the import of records from one CSV file. The order of the inserts does not matter.

For ordered bulk writers, order compliance is important to keep. In this case, only the same consecutive operations are grouped.

Currently, all bulk writes are optimized in memory. This is unfavorable for large bulk writes. In this case, one can use streaming bulk writes that only have a certain set of group operation in memory and when the maximum number of operations has been reached, operations are written to the database. The size can be specified.

Using ordered bulk writes. In this example we first insert some dog's name, add an attribute kind and change all dogs to cats. After that we delete three cats. This example would not work with unordered bulk writes.

bulk = "bulk"
       |> OrderedBulk.new()
       |> OrderedBulk.insert_one(%{name: "Greta"})
       |> OrderedBulk.insert_one(%{name: "Tom"})
       |> OrderedBulk.insert_one(%{name: "Waldo"})
       |> OrderedBulk.update_one(%{name: "Greta"}, %{"$set": %{kind: "dog"}})
       |> OrderedBulk.update_one(%{name: "Tom"}, %{"$set": %{kind: "dog"}})
       |> OrderedBulk.update_one(%{name: "Waldo"}, %{"$set": %{kind: "dog"}})
       |> OrderedBulk.update_many(%{kind: "dog"}, %{"$set": %{kind: "cat"}})
       |> OrderedBulk.delete_one(%{kind: "cat"})
       |> OrderedBulk.delete_one(%{kind: "cat"})
       |> OrderedBulk.delete_one(%{kind: "cat"})

result = Mongo.BulkWrite.write(:mongo, bulk, w: 1)

In the following example we import 1.000.000 integers into the MongoDB using the stream api:

We need to create an insert operation for each number. Then we call the Mongo.UnorderedBulk.stream function to import it. This function returns a stream function which accumulate all inserts operations until the limit 1000 is reached. In this case the operation group is send to MongoDB. So using the stream api you can reduce the memory using while importing big volume of data.

1..1_000_000 
|> Stream.map(fn i -> Mongo.BulkOps.get_insert_one(%{number: i}) end) 
|> Mongo.UnorderedBulk.write(:mongo, "bulk", 1_000)
|> Stream.run()

For more information see:

and have a look at the test units as well.

GridFS

The driver supports the GridFS specifications. You create a Mongo.GridFs.Bucket struct and with this struct you can upload and download files. For example:

    bucket = Bucket.new(top)
    upload_stream = Upload.open_upload_stream(bucket, "test.jpg")
    src_filename = "./test/data/test.jpg"
    File.stream!(src_filename, [], 512) |> Stream.into(upload_stream) |> Stream.run()

    file_id = upload_stream.id

In the example a new bucket with default values is used to upload a file from the file system (./test/data/test.jpg) to the MongoDB (using the name test.jpg). The upload_stream struct contains the id of the new file which can be used to download the stored file. The following code fragments downloads the file by using the file_id.

    dest_filename = "/tmp/my-test-file.jps"

    with {:ok, stream} <- Mongo.GridFs.Download.open_download_stream(bucket, file_id) do
      stream
      |> Stream.into(File.stream!(dest_filename))
      |> Stream.run
    end

For more information see:

Transactions

Since MongoDB 4.x, transactions for multiple write operations are possible. The Mongo.Session is responsible for the details and you can use a convenient api for transactions:

alias Mongo.Session

{:ok, ids} = Session.with_transaction(top, fn opts ->
{:ok, %InsertOneResult{:inserted_id => id1}} = Mongo.insert_one(top, "dogs", %{name: "Greta"}, opts)
{:ok, %InsertOneResult{:inserted_id => id2}} = Mongo.insert_one(top, "dogs", %{name: "Waldo"}, opts)
{:ok, %InsertOneResult{:inserted_id => id3}} = Mongo.insert_one(top, "dogs", %{name: "Tom"}, opts)
{:ok, [id1, id2, id3]}
end, w: 1)

It is also possible to get more control over the progress of the transaction:

alias Mongo.Session

{:ok, session} = Session.start_session(top, :write, [])
:ok = Session.start_transaction(session)

Mongo.insert_one(top, "dogs", %{name: "Greta"}, session: session)
Mongo.insert_one(top, "dogs", %{name: "Waldo"}, session: session)
Mongo.insert_one(top, "dogs", %{name: "Tom"}, session: session)

:ok = Session.commit_transaction(session)
:ok = Session.end_session(top, session)

For more information see:

Mongo.Session

and have a look at the test units as well.

Command Monitoring

You can watch all events that are triggered while the driver send requests and processes responses. You can use the Mongo.EventHandler as a starting point. It logs the events from the topic :commands (by ignoring the :isMaster command) to Logger.info:

iex> Mongo.EventHandler.start()
iex> {:ok, conn} = Mongo.start_link(url: "mongodb://localhost:27017/test")
{:ok, #PID<0.226.0>}
iex> Mongo.find_one(conn, "test", %{})
[info] Received command: %Mongo.Events.CommandStartedEvent{command: [find: "test", ...
[info] Received command: %Mongo.Events.CommandSucceededEvent{command_name: :find, ...

Testing and Travis CI

The travis.yml file uses only the latest MongoDB. It creates a replica set of three nodes and disables the SSL test case. If you want to run the test cases against other MongoDB deployments or older versions, you can use the mtools for deployment and run the test cases locally:

pyenv global 3.6
pip3 install --upgrade pip
pip3 install mtools[all]
export PATH=to-your-mongodb/bin/:$PATH
mlaunch init --setParameter enableTestCommands=1 --replicaset --name "rs_1"
mix test --exclude ssl --exclude socket

The SSL test suite is disabled by default.

Enable the SSL Tests

mix test --exclude ssl

Enable SSL on Your MongoDB Server

$ openssl req -newkey rsa:2048 -new -x509 -days 365 -nodes -out mongodb-cert.crt -keyout mongodb-cert.key
$ cat mongodb-cert.key mongodb-cert.crt > mongodb.pem
$ mongod --sslMode allowSSL --sslPEMKeyFile /path/to/mongodb.pem

For --sslMode you can use one of allowSSL or preferSSL
You can enable any other options you want when starting mongod

More Examples

There are some basic examples in the example folder. But if you want to see the driver in action take a look at Vega, especially the Board.ex module for using the transaction api together with bulk operations.

Special Thanks

Special thanks to JetBrains for providing a free JetBrains Open Source license for their complete toolbox.

The Documentation is online, but currently not up to date. This will be done as soon as possible. In the meantime, look in the source code. Especially for the individual options.

This driver is based on the original Elixir driver for MongoDB.

License

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Name		Name	Last commit message	Last commit date
Latest commit History 816 Commits
config		config
examples		examples
lib		lib
test		test
.ebert.yml		.ebert.yml
.gitignore		.gitignore
.travis.yml		.travis.yml
CHANGELOG.md		CHANGELOG.md
LICENSE		LICENSE
README.md		README.md
coveralls.json		coveralls.json
jetbrains.svg		jetbrains.svg
mix.exs		mix.exs
mix.lock		mix.lock

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

An Alternative Elixir Driver for MongoDB

Features

Data Representation

Usage

Installation

Simple Connection to MongoDB

Connection Pooling

Replica Sets

Auth Mechanisms

AWS, TLS and Erlang SSL Ciphers

Find

Change Streams

Inserts

Indexes

Bulk Writes

GridFS

Transactions

Command Monitoring

Testing and Travis CI

Enable the SSL Tests

Enable SSL on Your MongoDB Server

More Examples

Special Thanks

License

About

Releases

Packages

Languages

License

ben-dyer/elixir-mongodb-driver

Folders and files

Latest commit

History

Repository files navigation

An Alternative Elixir Driver for MongoDB

Features

Data Representation

Usage

Installation

Simple Connection to MongoDB

Connection Pooling

Replica Sets

Auth Mechanisms

AWS, TLS and Erlang SSL Ciphers

Find

Change Streams

Inserts

Indexes

Bulk Writes

GridFS

Transactions

Command Monitoring

Testing and Travis CI

Enable the SSL Tests

Enable SSL on Your MongoDB Server

More Examples

Special Thanks

License

About

Resources

License

Stars

Watchers

Forks

Releases

Packages 0

Languages

Packages