Graphene Core

Build Status:

master	develop	hardfork	testnet	bitshares-fc

• Getting Started
• Support
• Using the API
• Accessing restricted API's
• FAQ
• License

Graphene Core is the Graphene blockchain implementation and command-line interface. The web wallet is Graphene UI.

Visit gph.ai to learn about Graphene and join the community at forum.gph.ai.

Information for developers can be found in the Graphene Developer Portal. Users interested in how bitshares works can go to the Graphene Documentation site.

Getting Started

Build instructions and additional documentation are available in the wiki.

We recommend building on Ubuntu 16.04 LTS (64-bit)

Build Dependencies:

sudo apt-get update
sudo apt-get install autoconf cmake make automake libtool git libboost-all-dev libssl-dev g++ libcurl4-openssl-dev

Build Script:

git clone https://github.com/graphene-blockchain/graphene-core.git
cd graphene-core
git checkout graphene # may substitute "graphene" with current release tag
git submodule update --init --recursive
cmake -DCMAKE_BUILD_TYPE=RelWithDebInfo .
make

Upgrade Script (prepend to the Build Script above if you built a prior release):

git remote set-url origin https://github.com/graphene-blockchain/graphene-core.git
git checkout graphene
git remote set-head origin --auto
git pull
git submodule update --init --recursive # this command may fail
git submodule sync --recursive
git submodule update --init --recursive

NOTE: Versions of Boost 1.57 through 1.69 are supported. Newer versions may work, but have not been tested. If your system came pre-installed with a version of Boost that you do not wish to use, you may manually build your preferred version and use it with Graphene by specifying it on the CMake command line.

Example: cmake -DBOOST_ROOT=/path/to/boost .

NOTE: Graphene requires a 64-bit operating system to build, and will not build on a 32-bit OS.

NOTE: Graphene now supports Ubuntu 18.04 LTS

NOTE: Graphene now supports OpenSSL 1.1.0

After Building, the witness_node can be launched with:

./programs/witness_node/witness_node

The node will automatically create a data directory including a config file. It may take several hours to fully synchronize the blockchain. After syncing, you can exit the node using Ctrl+C and setup the command-line wallet by editing witness_node_data_dir/config.ini as follows:

rpc-endpoint = 127.0.0.1:8090

IMPORTANT: By default the witness node will start in reduced memory mode by using some of the commands detailed in Memory reduction for nodes. In order to run a full node with all the account history you need to remove partial-operations and max-ops-per-account from your config file. Please note that currently(2018-10-17) a full node will need more than 160GB of RAM to operate and required memory is growing fast. Consider the following table as minimal requirements before running a node:

Default	Full	Minimal	ElasticSearch
40G HDD, 8G RAM	80G SSD, 16G RAM *	40G HDD, 8G RAM	100G SSD, 16G RAM

* For this setup, allocate at least 500GB of SSD as swap.

After starting the witness node again, in a separate terminal you can run:

./programs/cli_wallet/cli_wallet

Set your inital password:

>>> set_password <PASSWORD>
>>> unlock <PASSWORD>

To import your initial balance:

>>> import_balance <ACCOUNT NAME> [<WIF_KEY>] true

If you send private keys over this connection, rpc-endpoint should be bound to localhost for security.

Use help to see all available wallet commands. Source definition and listing of all commands is available here.

Support

Technical support is available in the Graphene Forum technical support subforum.

Graphene Core bugs can be reported directly to the issue tracker.

Graphene UI bugs should be reported to the UI issue tracker

Using the API

We provide several different API's. Each API has its own ID. When running witness_node, initially two API's are available: API 0 provides read-only access to the database, while API 1 is used to login and gain access to additional, restricted API's.

Here is an example using wscat package from npm for websockets:

$ npm install -g wscat
$ wscat -c ws://127.0.0.1:8090
> {"id":1, "method":"call", "params":[0,"get_accounts",[["1.2.0"]]]}
< {"id":1,"result":[{"id":"1.2.0","annotations":[],"membership_expiration_date":"1969-12-31T23:59:59","registrar":"1.2.0","referrer":"1.2.0","lifetime_referrer":"1.2.0","network_fee_percentage":2000,"lifetime_referrer_fee_percentage":8000,"referrer_rewards_percentage":0,"name":"committee-account","owner":{"weight_threshold":1,"account_auths":[],"key_auths":[],"address_auths":[]},"active":{"weight_threshold":6,"account_auths":[["1.2.5",1],["1.2.6",1],["1.2.7",1],["1.2.8",1],["1.2.9",1],["1.2.10",1],["1.2.11",1],["1.2.12",1],["1.2.13",1],["1.2.14",1]],"key_auths":[],"address_auths":[]},"options":{"memo_key":"GPH1111111111111111111111111111111114T1Anm","voting_account":"1.2.0","num_witness":0,"num_committee":0,"votes":[],"extensions":[]},"statistics":"2.7.0","whitelisting_accounts":[],"blacklisting_accounts":[]}]}

We can do the same thing using an HTTP client such as curl for API's which do not require login or other session state:

$ curl --data '{"jsonrpc": "2.0", "method": "call", "params": [0, "get_accounts", [["1.2.0"]]], "id": 1}' http://127.0.0.1:8090/rpc
{"id":1,"result":[{"id":"1.2.0","annotations":[],"membership_expiration_date":"1969-12-31T23:59:59","registrar":"1.2.0","referrer":"1.2.0","lifetime_referrer":"1.2.0","network_fee_percentage":2000,"lifetime_referrer_fee_percentage":8000,"referrer_rewards_percentage":0,"name":"committee-account","owner":{"weight_threshold":1,"account_auths":[],"key_auths":[],"address_auths":[]},"active":{"weight_threshold":6,"account_auths":[["1.2.5",1],["1.2.6",1],["1.2.7",1],["1.2.8",1],["1.2.9",1],["1.2.10",1],["1.2.11",1],["1.2.12",1],["1.2.13",1],["1.2.14",1]],"key_auths":[],"address_auths":[]},"options":{"memo_key":"GPH1111111111111111111111111111111114T1Anm","voting_account":"1.2.0","num_witness":0,"num_committee":0,"votes":[],"extensions":[]},"statistics":"2.7.0","whitelisting_accounts":[],"blacklisting_accounts":[]}]}

API 0 is accessible using regular JSON-RPC:

$ curl --data '{"jsonrpc": "2.0", "method": "get_accounts", "params": [["1.2.0"]], "id": 1}' http://127.0.0.1:8090/rpc

Accessing restricted API's

You can restrict API's to particular users by specifying an api-access file in config.ini or by using the --api-access /full/path/to/api-access.json startup node command. Here is an example api-access file which allows user bytemaster with password supersecret to access four different API's, while allowing any other user to access the three public API's necessary to use the wallet:

{
   "permission_map" :
   [
      [
         "bytemaster",
         {
            "password_hash_b64" : "9e9GF7ooXVb9k4BoSfNIPTelXeGOZ5DrgOYMj94elaY=",
            "password_salt_b64" : "INDdM6iCi/8=",
            "allowed_apis" : ["database_api", "network_broadcast_api", "history_api", "network_node_api"]
         }
      ],
      [
         "*",
         {
            "password_hash_b64" : "*",
            "password_salt_b64" : "*",
            "allowed_apis" : ["database_api", "network_broadcast_api", "history_api"]
         }
      ]
   ]
}

Passwords are stored in base64 as salted sha256 hashes. A simple Python script, saltpass.py is avaliable to obtain hash and salt values from a password. A single asterisk "*" may be specified as username or password hash to accept any value.

With the above configuration, here is an example of how to call add_node from the network_node API:

{"id":1, "method":"call", "params":[1,"login",["bytemaster", "supersecret"]]}
{"id":2, "method":"call", "params":[1,"network_node",[]]}
{"id":3, "method":"call", "params":[2,"add_node",["127.0.0.1:9090"]]}

Note, the call to network_node is necessary to obtain the correct API identifier for the network API. It is not guaranteed that the network API identifier will always be 2.

Since the network_node API requires login, it is only accessible over the websocket RPC. Our doxygen documentation contains the most up-to-date information about API's for the witness node and the wallet. If you want information which is not available from an API, it might be available from the database; it is fairly simple to write API methods to expose database methods.

FAQ

• Is there a way to generate help with parameter names and method descriptions?

  Yes. Documentation of the code base, including APIs, can be generated using Doxygen. Simply run doxygen in this directory.

  If both Doxygen and perl are available in your build environment, the CLI wallet's help and gethelp commands will display help generated from the doxygen documentation.

  If your CLI wallet's help command displays descriptions without parameter names like signed_transaction transfer(string, string, string, string, string, bool) it means CMake was unable to find Doxygen or perl during configuration. If found, the output should look like this: signed_transaction transfer(string from, string to, string amount, string asset_symbol, string memo, bool broadcast)

• Is there a way to allow external program to drive cli_wallet via websocket, JSONRPC, or HTTP?

  Yes. External programs may connect to the CLI wallet and make its calls over a websockets API. To do this, run the wallet in server mode, i.e. cli_wallet -s "127.0.0.1:9999" and then have the external program connect to it over the specified port (in this example, port 9999).

• Is there a way to access methods which require login over HTTP?

  No. Login is inherently a stateful process (logging in changes what the server will do for certain requests, that's kind of the point of having it). If you need to track state across HTTP RPC calls, you must maintain a session across multiple connections. This is a famous source of security vulnerabilities for HTTP applications. Additionally, HTTP is not really designed for "server push" notifications, and we would have to figure out a way to queue notifications for a polling client.

  Websockets solves all these problems. If you need to access Graphene's stateful methods, you need to use Websockets.

• What is the meaning of a.b.c numbers?

  The first number specifies the space. Space 1 is for protocol objects, 2 is for implementation objects. Protocol space objects can appear on the wire, for example in the binary form of transactions. Implementation space objects cannot appear on the wire and solely exist for implementation purposes, such as optimization or internal bookkeeping.

  The second number specifies the type. The type of the object determines what fields it has. For a complete list of type ID's, see enum object_type and enum impl_object_type in bitshares/libraries/chain/include/graphene/chain/protocol/types.hpp

  The third number specifies the instance. The instance of the object is different for each individual object.

• The answer to the previous question was really confusing. Can you make it clearer?

  All account ID's are of the form 1.2.x. If you were the 9735th account to be registered, your account's ID will be 1.2.9735. Account 0 is special (it's the "committee account," which is controlled by the committee members and has a few abilities and restrictions other accounts do not).

  All asset ID's are of the form 1.3.x. If you were the 29th asset to be registered, your asset's ID will be 1.3.29. Asset 0 is special (it's BTS, which is considered the "core asset").

  The first and second number together identify the kind of thing you're talking about (1.2 for accounts, 1.3 for assets). The third number identifies the particular thing.

• How do I get the network_add_nodes command to work? Why is it so complicated?

  You need to follow the instructions in the "Accessing restricted API's" section to allow a username/password access to the network_node API. Then you need to pass the username/password to the cli_wallet on the command line or in a config file.

  It's set up this way so that the default configuration is secure even if the RPC port is publicly accessible. It's fine if your witness_node allows the general public to query the database or broadcast transactions (in fact, this is how the hosted web UI works). It's less fine if your witness_node allows the general public to control which p2p nodes it's connecting to. Therefore the API to add p2p connections needs to be set up with proper access controls.

License

Graphene Core is under the MIT license. See LICENSE for more information.