bigchaindb/docs/source/nodes/setup-run-node.md

Set Up and Run a Node

This section goes through the steps to set up a BigchainDB node (running RethinkDB Server, BigchainDB Server, etc.). There are instructions for two cases:

1. Stand-Alone Node (useful for local development)
2. Cluster Node

Check the Node Requirements

The first step is to make sure you have a server (or equivalent) which meets the requirements for a BigchainDB node.

System Clock Sync

If you're just setting up a stand-alone node, then you can skip this step.

A BigchainDB node uses its system clock to generate timestamps for blocks and votes, so that clock should be kept in sync with some standard clock(s). The standard way to do that is to run an NTP daemon (Network Time Protocol daemon) on the node. (You could also use tlsdate, which uses TLS timestamps rather than NTP, but don't: it's not very accurate and it will break with TLS 1.3, which removes the timestamp.)

NTP is a standard protocol. There are many NTP daemons implementing it. We don't recommend a particular one. On the contrary, we recommend that different nodes in a federation run different NTP daemons, so that a problem with one daemon won't affect all nodes. Some options are:

• The reference NTP daemon from ntp.org; see their support website
• OpenNTPD
• chrony
• Maybe NTPsec, once it's production-ready
• Maybe Ntimed, once it's production-ready
• More

It's tricky to make an NTP daemon setup secure. Always install the latest version and read the documentation about how to configure and run it securely.

The appendix has some notes on NTP daemon setup.

Set Up the File System for RethinkDB

If you're just setting up a stand-alone node, then you can probably skip this step.

Ideally, use a file system that supports direct I/O (Input/Output), a feature whereby file reads and writes go directly from RethinkDB to the storage device, bypassing the operating system read and write caches.

TODO: What file systems support direct I/O? How can you check? How do you enable it, if necessary?

See def install_rethinkdb() in deploy-cluster-aws/fabfile.py for an example of configuring a file system on an AWS instance running Ubuntu.

Mount the partition for RethinkDB on /data: we will tell RethinkDB to store its data there.

TODO: This section needs more elaboration

Install RethinkDB Server

If you don't already have RethinkDB Server installed, you must install it. The RethinkDB documentation has instructions for how to install RethinkDB Server on a variety of operating systems.

Configure RethinkDB Server

Stand-Alone Node

If you're setting up a stand-alone node (i.e. not intending for it to join a cluster), then the default RethinkDB configuration will work.

Cluster Node

Create a RethinkDB configuration file (text file) named instance.conf in the /etc/rethinkdb/instances.d/ directory with the following contents (explained below):

directory=/data
bind=all
direct-io
# Replace node?_hostname with actual node hostnames below, e.g. rdb.examples.com
join=node0_hostname:29015
join=node1_hostname:29015
join=node2_hostname:29015
# continue until there's a join= line for each node in the federation

• directory=/data tells the RethinkDB node to store its share of the database data in /data.
• bind=all binds RethinkDB to all local network interfaces (e.g. loopback, Ethernet, wireless, whatever is available), so it can communicate with the outside world. (The default is to bind only to local interfaces.)
• direct-io tells RethinkDB to use direct I/O (explained earlier).
• join=hostname:29015 lines: A cluster node needs to find out the hostnames of all the other nodes somehow. You could designate one node to be the one that every other node asks, and put that node's hostname in the config file, but that wouldn't be very decentralized. Instead, we include every node in the list of nodes-to-ask.

If you're curious about the RethinkDB config file, there's a RethinkDB documentation page about it.

TODO: Explain how to configure the RethinkDB cluster to be more secure.

Run RethinkDB Server

You can run RethinkDB by opening a Terminal and entering rethinkdb. You could do that now or wait until just before you start BigchainDB.

Install Python 3.4+

If you don't already have it, then you should install Python 3.4+.

If you're testing or developing BigchainDB on a stand-alone node, then you should probably create a Python 3.4+ virtual environment and activate it (e.g. using virtualenv or conda). Later we will install several Python packages and you probably only want those installed in the virtual environment.

Install BigchainDB Server

BigchainDB Server has some OS-level dependencies that must be installed.

On Ubuntu 14.04, we found that the following was enough:

sudo apt-get update
sudo apt-get install g++ python3-dev

On Fedora 23, we found that the following was enough (tested in February 2015):

sudo dnf update
sudo dnf install gcc-c++ redhat-rpm-config python3-devel

(If you're using a version of Fedora before version 22, you may have to use yum instead of dnf.)

With OS-level dependencies installed, you can install BigchainDB Server with pip or from source.

How to Install BigchainDB with pip

BigchainDB (i.e. both the Server and the officially-supported drivers) is distributed as a Python package on PyPI so you can install it using pip. First, make sure you have a version of pip installed for Python 3.4+:

pip -V

If it says that pip isn't installed, or it says pip is associated with a Python version less than 3.4, then you must install a pip version associated with Python 3.4+. See the pip installation instructions. On Ubuntu 14.04, we found that this works:

sudo apt-get install python3-setuptools
sudo easy_install3 pip
pip3 install --upgrade pip wheel setuptools

(Note: Using sudo apt-get python3-pip also installs a Python 3 version of pip (named pip3) but we found it installed a very old version and there were issues with updating it.)

Once you have a version of pip associated with Python 3.4+, then you can install BigchainDB Server (and officially-supported BigchainDB drivers) using:

sudo pip install bigchaindb

(or maybe sudo pip3 install bigchaindb or sudo pip3.4 install bigchaindb. The sudo may not be necessary.)

Note: You can use pip to upgrade the bigchaindb package to the latest version using sudo pip install --upgrade bigchaindb

How to Install BigchainDB from Source

If you want to install BitchainDB from source because you want to contribute code (i.e. as a BigchainDB developer), then please see the instructions in the CONTRIBUTING.md file.

Otherwise, clone the public repository:

git clone git@github.com:bigchaindb/bigchaindb.git

and then install from source:

python setup.py install

Configure BigchainDB Server

Start by creating a default BigchainDB configuration file (named .bigchaindb) in your $HOME directory using:

bigchaindb -y configure

There's documentation for the bigchaindb command is in the section on the BigchainDB Command Line Interface (CLI).

Stand-Alone Node

The default BigchainDB configuration file will work.

Cluster Node

Open $HOME/.bigchaindb in your text editor and:

• Change "server": {"bind": "localhost:9984", ... } to "server": {"bind": "0.0.0.0:9984", ... }. This makes it so traffic can come from any IP address to port 9984 (the HTTP Client-Server API port).
• Change "api_endpoint": "http://localhost:9984/api/v1" to "api_endpoint": "http://your_api_hostname:9984/api/v1"
• Change "keyring": [] to "keyring": ["public_key_of_other_node_A", "public_key_of_other_node_B", "..."] i.e. a list of the public keys of all the other nodes in the federation. The keyring should not include your node's public key.

For more information about the BigchainDB config file, see Configuring a BigchainDB Node.

Run BigchainDB Server

Once you've installed BigchainDB Server, you can run it. First make sure you have RethinkDB running:

rethinkdb

You can verify that RethinkDB is running by opening the RethinkDB web interface in your web browser. It should be at http://rethinkdb-hostname:8080/. If you're running RethinkDB on localhost, that would be http://localhost:8080/.

You can start BigchainDB Server using:

bigchaindb start

If it's the first time you've run bigchaindb start, then it creates the database (a RethinkDB database), the tables, the indexes, and the genesis block. It then starts BigchainDB. If you're run bigchaindb start or bigchaindb init before (and you haven't dropped the database), then bigchaindb start just starts BigchainDB.

Run BigchainDB with Docker

NOT for Production Use

For those who like using Docker and wish to experiment with BigchainDB in non-production environments, we currently maintain a Docker image and a Dockerfile that can be used to build an image for bigchaindb.

Pull and Run the Image from Docker Hub

Assuming you have Docker installed, you would proceed as follows.

In a terminal shell, pull the latest version of the BigchainDB Docker image using:

docker pull bigchaindb/bigchaindb

then do a one-time configuration step to create the config file; we will use the -y option to accept all the default values. The configuration file will be stored in a file on your host machine at ~/bigchaindb_docker/.bigchaindb:

docker run --rm -v "$HOME/bigchaindb_docker:/data" -ti \
  bigchaindb/bigchaindb -y configure
Generating keypair
Configuration written to /data/.bigchaindb
Ready to go!

Let's analyze that command:

• docker run tells Docker to run some image
• --rm remove the container once we are done
• -v "$HOME/bigchaindb_docker:/data" map the host directory $HOME/bigchaindb_docker to the container directory /data; this allows us to have the data persisted on the host machine, you can read more in the official Docker documentation
• -t allocate a pseudo-TTY
• -i keep STDIN open even if not attached
• `bigchaindb/bigchaindb the image to use
• -y configure execute the configure sub-command (of the bigchaindb command) inside the container, with the -y option to automatically use all the default config values

After configuring the system, you can run BigchainDB with the following command:

docker run -v "$HOME/bigchaindb_docker:/data" -d \
  --name bigchaindb \
  -p "58080:8080" -p "59984:9984" \
  bigchaindb/bigchaindb start

The command is slightly different from the previous one, the differences are:

• -d run the container in the background
• --name bigchaindb give a nice name to the container so it's easier to refer to it later
• -p "58080:8080" map the host port 58080 to the container port 8080 (the RethinkDB admin interface)
• -p "59984:9984" map the host port 59984 to the container port 9984 (the BigchainDB API server)
• start start the BigchainDB service

Another way to publish the ports exposed by the container is to use the -P (or --publish-all) option. This will publish all exposed ports to random ports. You can always run docker ps to check the random mapping.

You can also access the RethinkDB dashboard at http://localhost:58080/

If that doesn't work, then replace localhost with the IP or hostname of the machine running the Docker engine. If you are running docker-machine (e.g. on Mac OS X) this will be the IP of the Docker machine (docker-machine ip machine_name).

Load Testing with Docker

Now that we have BigchainDB running in the Docker container named bigchaindb, we can start another BigchainDB container to generate a load test for it.

First, make sure the container named bigchaindb is still running. You can check that using:

docker ps

You should see a container named bigchaindb in the list.

You can load test the BigchainDB running in that container by running the bigchaindb load command in a second container:

docker run --rm -v "$HOME/bigchaindb_docker:/data" -ti \
  --link bigchaindb \
  bigchaindb/bigchaindb load

Note the --link option to link to the first container (named bigchaindb).

Aside: The bigchaindb load command has several options (e.g. -m). You can read more about it in the documentation about the BigchainDB command line interface.

If you look at the RethinkDB dashboard (in your web browser), you should see the effects of the load test. You can also see some effects in the Docker logs using:

docker logs -f bigchaindb

Building Your Own Image

Assuming you have Docker installed, you would proceed as follows.

In a terminal shell:

git clone git@github.com:bigchaindb/bigchaindb.git

Build the Docker image:

docker build --tag local-bigchaindb .

Now you can use your own image to run BigchainDB containers.