Run Chainhook as a Service Using Bitcoind
You can run Chainhook as a service to evaluate your if_this / then_that
predicates against the Bitcoin blockchain, delivering data—either file appendations or HTTP POST requests to a server you designate—for your application's use case. You can also dynamically register new predicates as the service is running by enabling the predicates registration API.
Prerequisites
Setting up a Bitcoin Node
The Bitcoin Core daemon (bitcoind) is a program that implements the Bitcoin protocol for remote procedure call (RPC) use. Chainhook can be set up to interact with the Bitcoin chainstate through bitcoind's ZeroMQ interface, its embedded networking library, passing raw blockchain data to be evaluated for relevant events.
This guide is written to work with the latest Bitcoin Core software containing bitcoind, Bitcoin Core 25.0.
NOTE:
While bitcoind can and will start syncing a Bitcoin node, customizing this node to your use cases beyond supporting a Chainhook is out of scope for this guide. See the Bitcoin wiki for "Running Bitcoin" or bitcoin.org's Running A Full Node guide.
- Make note of the path of your
bitcoind
executable (located within thebin
directory of thebitcoin-25.0
folder you downloaded above appropriate to your operating system). - Navigate to your project folder where your Chainhook node will reside, create a new file, and rename it to
bitcoin.conf
. Copy the configuration below to thisbitcoin.conf
file. - Find and copy your Bitcoin data directory and paste to the
datadir
field in thebitcoin.conf
file below. Either copy the default path (see list of default directories by operating system) or copy the custom path you set for your Bitcoin data. - Set a username of your choice for bitcoind and use it in the
rpcuser
configuration below (devnet
is a default). - Set a password of your choice for bitcoind and use it in the
rpcpassword
configuration below (devnet
is a default).
# Bitcoin Core Configuration
datadir=/path/to/bitcoin/directory/ # Path to Bitcoin directory
server=1
rpcuser=devnet
rpcpassword=devnet
rpcport=8332
rpcallowip=0.0.0.0/0
rpcallowip=::/0
txindex=1
listen=1
discover=0
dns=0
dnsseed=0
listenonion=0
rpcserialversion=1
disablewallet=0
fallbackfee=0.00001
rpcthreads=8
blocksonly=1
dbcache=4096
# Start zeromq
zmqpubhashblock=tcp://0.0.0.0:18543
NOTE:
The below command is a startup process that, if this is your first time syncing a node, might take a few hours to a few days to run. Alternatively, if the directory pointed to in the
datadir
field above contains bitcoin blockchain data, syncing will resume.
Now that you have the bitcoin.conf
file ready with the bitcoind configurations, you can run the bitcoind node. The command takes the form path/to/bitcoind -conf=path/to/bitcoin.conf
, for example:
/Volumes/SSD/bitcoin-25.0/bin/bitcoind -conf=/Volumes/SSD/project/Chainhook/bitcoin.conf
Once the above command runs, you will see zmq_url
entries in the console's stdout, displaying ZeroMQ logs of your bitcoin node.
Configure Chainhook
In this section, you will configure Chainhook to match the network configurations with the bitcoin config file. First, install the latest version of Chainhook.
Next, you will generate a Chainhook.toml
file to connect Chainhook with your bitcoind node. Navigate to the directory where you want to generate the Chainhook.toml
file and use the following command in your terminal:
chainhook config generate --mainnet
Several network parameters in the generated Chainhook.toml
configuration file need to match those in the bitcoin.conf
file created earlier in the Setting up a Bitcoin Node section. Please update the following parameters accordingly:
- Update
bitcoind_rpc_username
with the username set forrpcuser
inbitcoin.conf
. - Update
bitcoind_rpc_password
with the password set forrpcpassword
inbitcoin.conf
. - Update
bitcoind_rpc_url
with the same host and port used forrpcport
inbitcoin.conf
.
Additionally, if you want to receive events from the configured Bitcoin node, substitute stacks_node_rpc_url
with bitcoind_zmq_url
, as follows:
[storage]
working_dir = "cache"
# The Http Api allows you to register / deregister
# predicates dynamically.
# This is disabled by default.
# [http_api]
# http_port = 20456
# database_uri = "redis://localhost:6379/"
[network]
mode = "mainnet"
bitcoind_rpc_url = "http://localhost:8332"
bitcoind_rpc_username = "devnet"
bitcoind_rpc_password = "devnet"
# Bitcoin block events can be received by Chainhook
# either through a Bitcoin node's ZeroMQ interface,
# or through the Stacks node. The Stacks node is
# used by default:
# stacks_node_rpc_url = "http://localhost:20443"
# but zmq can be used instead:
bitcoind_zmq_url = "tcp://0.0.0.0:18543"
[limits]
max_number_of_bitcoin_predicates = 100
max_number_of_concurrent_bitcoin_scans = 100
max_number_of_stacks_predicates = 10
max_number_of_concurrent_stacks_scans = 10
max_number_of_processing_threads = 16
max_number_of_networking_threads = 16
max_caching_memory_size_mb = 32000
[[event_source]]
tsv_file_url = "https://archive.hiro.so/mainnet/stacks-blockchain-api/mainnet-stacks-blockchain-api-latest"
Here is a table of the relevant parameters this guide changes in our configuration files.
bitcoin.conf | Chainhook.toml |
---|---|
rpcuser | bitcoind_rpc_username |
rpcpassword | bitcoind_rpc_password |
rpcport | bitcoind_rpc_url |
zmqpubhashblock | bitcoind_zmq_url |
Scan blockchain based on predicates
Now that your bitcoind and Chainhook configurations are complete, you can define the Chainhook predicates you would like to scan against bitcoin blocks. These predicates are where you specify the kind of blockchain events that trigger Chainhook to deliver a result (either a file appendation or an HTTP POST request). This section helps you with an example JSON file to scan a range of blocks in the blockchain to trigger results. To understand the supported predicates for Bitcoin, refer to how to use chainhooks with bitcoin.
The following is an example to walk you through an if_this / then_that
predicate design that appends event payloads to the configured file destination.
Example 1 - file_append
To generate a sample JSON file with predicates, execute the following command in your terminal:
chainhook predicates new stacking-pool.json --bitcoin
Replace the contents of the stacking-pool.json
file with the following:
{
"chain": "bitcoin",
"uuid": "1",
"name": "Stacking Pool",
"version": 1,
"networks": {
"mainnet": {
"start_block": 801500,
"end_block": 802000,
"if_this": {
"scope": "outputs",
"p2wpkh": {
"equals": "bc1qs0kkdpsrzh3ngqgth7mkavlwlzr7lms2zv3wxe"
}
},
"then_that": {
"file_append": {
"path": "bitcoin-transactions.txt"
}
}
}
}
}
This example demonstrates scanning a portion of the Bitcoin blockchain to capture specific outputs from a Bitcoin address associated with a Stacking pool, Friedgar Pool.
NOTE:
You can get blockchain height and current block by referring to https://explorer.hiro.so/blocks?chain=mainnet
Now, use the following command to scan the blocks based on the predicates defined in the stacking-pool.json
file.
chainhook predicates scan stacking-pool.json --config-path=./Chainhook.toml
The output of the above command will be a text file bitcoin-transactions.txt
generated based on the predicate definition.
Example 2 - http_post
Let's generate another sample predicate, this time we are going to send the payload to an API endpoint:
chainhook predicates new stacking-pool-api.json --bitcoin
Replace the contents of the stacking-pool-api.json
file with the following:
{
"chain": "bitcoin",
"uuid": "2",
"name": "Stacking Pool (API)",
"version": 1,
"networks": {
"mainnet": {
"start_block": 801500,
"if_this": {
"scope": "outputs",
"p2wpkh": {
"equals": "bc1qs0kkdpsrzh3ngqgth7mkavlwlzr7lms2zv3wxe"
}
},
"then_that": {
"http_post": {
"url": "http://localhost:3000/events",
"authorization_header": "12345"
}
}
}
}
}
NOTE:
The
start_block
is a required field when using thehttp_post
then_that
predicate.
Once you are finished setting up your endpoint, use the following command to scan the blocks based on the predicates defined in the stacking-pool-api.json
file.
chainhook predicates scan stacking-pool-api.json --config-path=./Chainhook.toml
The above command posts events to the URL, http://localhost:3000/events
mentioned in the JSON file.
Initiate Chainhook Service
In the examples above, our Chainhook scanned historical blockchain data against predicates and delivered results. In this next section, let's learn how to set up a Chainhook that acts as an ongoing observer and event-streaming service.
We can start a Chainhook service with an existing predicate. We can also dynamically register new predicates by making an API call to our Chainhook. In both of these instances, our predicates will be delivering their results to a server set up to receive results.
- Initiate the chainhook service by passing the predicate path to the command as shown below:
chainhook service start --predicate-path=stacking-pool-api.json --config-path=Chainhook.toml
The above command registers the predicate based on the predicate definition in the stacking-pool-api.json
file.
Dynamically Register Predicates
You can also dynamically register new predicates with your Chainhook service.
First, we need to uncomment the following lines of code in the Chainhook.toml
file to enable the predicate registration server:
# ...
[http_api]
http_port = 20456
database_uri = "redis://localhost:6379/"
# ...
NOTE:
This assumes you have a local instance of Redis running.
Start the Chainhook service by running the following command:
chainhook service start --config-path=Chainhook.toml
To dynamically register a new predicate, send a POST request to the running predicate registration server at localhost:20456/v1/chainhooks
. Include the new predicate in JSON format within the request body. Use the following curl
command as an example:
curl -X POST \
-H "Content-Type: application/json" \
-d '{
"chain": "bitcoin",
"uuid": "3",
"name": "Ordinals",
"version": 1,
"networks": {
"mainnet": {
"start_block": 777534,
"if_this": {
"scope": "ordinals_protocol",
"operation": "inscription_feed"
},
"then_that": {
"http_post": {
"url": "http://localhost:3000/events",
"authorization_header": "12345"
}
}
}
}
}' \
http://localhost:20456/v1/chainhooks
The sample response should look like this:
{
"chainhook": {
"predicate": {
"operation": "inscription_feed",
"scope": "ordinals_protocol"
},
"uuid": "1"
},
"apply": [
{
"block_identifier": {
"hash": "0x00000000000000000003e3e2ffd3baaff2cddda7d12e84ed0ffe6f7778e988d4",
"index": 777534
},
"metadata": {},
"parent_block_identifier": {
"hash": "0x0000000000000000000463a1034c59e6dc94c7e52855582af11882743b86e2a7",
"index": 777533
},
"timestamp": 1676923039,
"transactions": [
{
"transaction_identifier": {
"hash": "0xca20efe5e4d71c16cd9b8dfe4d969efdd225ef0a26136a6a4409cb3afb2e013e"
},
"metadata": {
"ordinal_operations": [
{
"inscription_revealed": {
"content_bytes": "<INSCRIPTION_BYTES>",
"content_length": 12293,
"content_type": "image/jpeg",
"inscriber_address": "bc1punnjva5ayg84kf5tmvx265uwvp8py3ux24skz43aycj5rzdgzjfq0jxsuc",
"inscription_fee": 64520,
"inscription_id": "ca20efe5e4d71c16cd9b8dfe4d969efdd225ef0a26136a6a4409cb3afb2e013ei0",
"inscription_number": 0,
"inscription_output_value": 10000,
"ordinal_block_height": 543164,
"ordinal_number": 1728956147664701,
"ordinal_offset": 1147664701,
"satpoint_post_inscription": "ca20efe5e4d71c16cd9b8dfe4d969efdd225ef0a26136a6a4409cb3afb2e013e:0:0",
"transfers_pre_inscription": 0
}
}
],
"proof": null
},
"operations": []
// Other transactions
}
]
}
],
"rollback": []
}
Understand the output of the above JSON file with the following details.
- The
apply
payload includes the block header and the transactions that triggered the predicate. - The
rollback
payload includes the block header and the transactions that triggered the predicate for a past block that is no longer part of the canonical chain and must be reverted. (Note: This is a chief component of Chainhook's reorg aware functionality, maintaining rollback data for blocks near the chaintip.)
TIP:
You can also run chainhook service by passing multiple predicates. Example:
chainhook service start --predicate-path=predicate_1.json --predicate-path=predicate_2.json --config-path=Chainhook.toml
References
- To learn more about Ordinals, refer to Introducing Ordinals Explorer and Ordinals API.
- The OpenAPI specification for chainhook is available to understand the scope of chainhook.