This document specifies common definitions and requirements affecting Engine API specification in general.
- Underlying protocol
- Versioning
- Message ordering
- Load-balancing and advanced configurations
- Errors
- Timeouts
- Encoding
- Capabilities
Message format and encoding notation used by this specification are inherited from Ethereum JSON-RPC Specification.
Client software MUST expose Engine API at a port independent from JSON-RPC API.
The default port for the Engine API is 8551.
The Engine API is exposed under the engine
namespace.
To facilitate an Engine API consumer to access state and logs (e.g. proof-of-stake deposits) through the same connection,
the client MUST also expose the following subset of eth
methods:
eth_blockNumber
eth_call
eth_chainId
eth_getCode
eth_getBlockByHash
eth_getBlockByNumber
eth_getLogs
eth_sendRawTransaction
eth_syncing
These methods are described in Ethereum JSON-RPC Specification.
Engine API uses JWT authentication enabled by default. JWT authentication is specified in Authentication document.
The versioning of the Engine API is defined as follows:
- The version of each method and structure is independent from versions of other methods and structures.
- The
VX
, where theX
is the number of the version, is suffixed to the name of each method and structure. - The version of a method or a structure MUST be incremented by one if any of the following is changed:
- a set of method parameters
- a method response value
- a method behavior
- a set of structure fields
- The specification MAY reference a method or a structure without the version suffix e.g.
engine_newPayload
. These statements should be read as related to all versions of the referenced method or structure.
Consensus Layer client software MUST respect the order of the corresponding fork choice update events
when making calls to the engine_forkchoiceUpdated
method.
Execution Layer client software MUST process engine_forkchoiceUpdated
method calls
in the same order as they have been received.
The Engine API supports a one-to-many Consensus Layer to Execution Layer configuration. Intuitively this is because the Consensus Layer drives the Execution Layer and thus can drive many of them independently.
On the other hand, generic many-to-one Consensus Layer to Execution Layer configurations are not supported out-of-the-box.
The Execution Layer, by default, only supports one chain head at a time and thus has undefined behavior when multiple Consensus Layers simultaneously control the head.
The Engine API does work properly, if in such a many-to-one configuration, only one Consensus Layer instantiation is able to write to the Execution Layer's chain head and initiate the payload build process (i.e. call engine_forkchoiceUpdated
),
while other Consensus Layers can only safely insert payloads (i.e. engine_newPayload
) and read from the Execution Layer.
The list of error codes introduced by this specification can be found below.
Code | Message | Meaning |
---|---|---|
-32700 | Parse error | Invalid JSON was received by the server. |
-32600 | Invalid Request | The JSON sent is not a valid Request object. |
-32601 | Method not found | The method does not exist / is not available. |
-32602 | Invalid params | Invalid method parameter(s). |
-32603 | Internal error | Internal JSON-RPC error. |
-32000 | Server error | Generic client error while processing request. |
-38001 | Unknown payload | Payload does not exist / is not available. |
-38002 | Invalid forkchoice state | Forkchoice state is invalid / inconsistent. |
-38003 | Invalid payload attributes | Payload attributes are invalid / inconsistent. |
-38004 | Too large request | Number of requested entities is too large. |
-38005 | Unsupported fork | Payload belongs to a fork that is not supported. |
Each error returns a null
data
value, except -32000
which returns the data
object with a err
member that explains the error encountered.
For example:
$ curl https://localhost:8551 \
-X POST \
-H "Content-Type: application/json" \
-d '{"jsonrpc":"2.0","method":"engine_getPayloadV1","params": ["0x1"],"id":1}'
{
"jsonrpc": "2.0",
"id": 1,
"error": {
"code": -32000,
"message": "Server error",
"data": {
"err": "Database corrupted"
}
}
}
Consensus Layer client software MUST wait for a specified timeout
before aborting the call. In such an event, the Consensus Layer client software SHOULD retry the call when it is needed to keep progressing.
Consensus Layer client software MAY wait for response longer than it is specified by the timeout
parameter.
Values of a field of DATA
type MUST be encoded as a hexadecimal string with a 0x
prefix matching the regular expression ^0x(?:[a-fA-F0-9]{2})*$
.
Values of a field of QUANTITY
type MUST be encoded as a hexadecimal string with a 0x
prefix and the leading 0s stripped (except for the case of encoding the value 0
) matching the regular expression ^0x(?:0|(?:[a-fA-F1-9][a-fA-F0-9]*))$
.
Note: Byte order of encoded value having QUANTITY
type is big-endian.
Execution and consensus layer client software may exchange with a list of supported Engine API methods by calling engine_exchangeCapabilities
method.
Execution layer clients MUST support engine_exchangeCapabilities
method, while consensus layer clients are free to choose whether to call it or not.
Note: The method itself doesn't have a version suffix.
- method:
engine_exchangeCapabilities
- params:
Array of string
-- Array of strings, each string is a name of a method supported by consensus layer client software.
- timeout: 1s
Array of string
-- Array of strings, each string is a name of a method supported by execution layer client software.
-
Consensus and execution layer client software MAY exchange with a list of currently supported Engine API methods. Execution layer client software MUST NOT log any error messages if this method has either never been called or hasn't been called for a significant amount of time.
-
Request and response lists MUST contain Engine API methods that are currently supported by consensus and execution client software respectively. Name of each method in both lists MUST include suffixed version. Consider the following examples:
- Client software of both layers currently supports
V1
andV2
versions ofengine_newPayload
method:- params:
[["engine_newPayloadV1", "engine_newPayloadV2", ...]]
, - response:
["engine_newPayloadV1", "engine_newPayloadV2", ...]
.
- params:
V1
method has been deprecated andV3
method has been introduced on execution layer side since the last call:- params:
[["engine_newPayloadV1", "engine_newPayloadV2", ...]]
, - response:
["engine_newPayloadV2", "engine_newPayloadV3", ...]
.
- params:
- The same capabilities modification has happened in consensus layer client, so, both clients have the same capability set again:
- params:
[["engine_newPayloadV2", "engine_newPayloadV3", ...]]
, - response:
["engine_newPayloadV2", "engine_newPayloadV3", ...]
.
- params:
- Client software of both layers currently supports
-
The
engine_exchangeCapabilities
method MUST NOT be returned in the response list.