Workaound LlamaNodes.com issues

eth_defi/chain.py

@@ -19,6 +19,7 @@
 
 from eth_defi.event_reader.conversion import convert_jsonrpc_value_to_int
 from eth_defi.middleware import http_retry_request_with_sleep_middleware
+from eth_defi.provider.llamarpc import is_llama_bad_grapql_reply
 from eth_defi.provider.named import NamedProvider
 
 #: List of chain ids that need to have proof-of-authority middleweare installed
@@ -161,6 +162,26 @@ def middleware(method: RPCEndpoint, params: Any) -> Optional[RPCResponse]:
     return api_counter
 
 
+def get_graphql_url(provider: BaseProvider) -> str:
+    """Resolve potential GraphQL endpoint API for a JSON-RPC provider.
+
+    See :py:func:`has_graphql_support`.
+    """
+
+    # See BaseNamedProvider
+    if hasattr(provider, "call_endpoint_uri"):
+        base_url = provider.call_endpoint_uri
+    elif hasattr(provider, "endpoint_uri"):
+        # HTTPProvider
+        base_url = provider.endpoint_uri
+    else:
+        raise AssertionError(f"Do not know how to extract endpoint URI: {provider}")
+
+    graphql_url = urljoin(base_url, "graphql")
+
+    return graphql_url
+
+
 def has_graphql_support(provider: BaseProvider) -> bool:
     """Check if a node has GoEthereum GraphQL API turned on.
 
@@ -179,19 +200,11 @@ def has_graphql_support(provider: BaseProvider) -> bool:
         {"data":{"block":{"number":16328259}}}
     """
 
-    # See BaseNamedProvider
-    if hasattr(provider, "call_endpoint_uri"):
-        base_url = provider.call_endpoint_uri
-    elif hasattr(provider, "endpoint_uri"):
-        # HTTPProvider
-        base_url = provider.endpoint_uri
-    else:
-        raise AssertionError(f"Do not know how to extract endpoint URI: {provider}")
+    graphql_url = get_graphql_url(provider)
 
-    graphql_url = urljoin(base_url, "graphql")
     try:
         resp = requests.get(graphql_url)
-        return resp.status_code == 400
+        return resp.status_code == 400 and not is_llama_bad_grapql_reply(resp)
     except Exception as e:
         # ConnectionError, etc.
         return False

eth_defi/event_reader/reorganisation_monitor.py

@@ -16,7 +16,7 @@
 from tqdm import tqdm
 from web3 import Web3, HTTPProvider
 
-from eth_defi.chain import has_graphql_support
+from eth_defi.chain import has_graphql_support, get_graphql_url
 from eth_defi.event_reader.block_header import BlockHeader, Timestamp
 from eth_defi.provider.fallback import FallbackProvider
 from eth_defi.provider.mev_blocker import MEVBlockerProvider
@@ -737,7 +737,8 @@ def create_reorganisation_monitor(web3: Web3, check_depth=250) -> Reorganisation
     if has_graphql_support(provider):
         # 10x faster /graphql implementation,
         # not provided by public nodes
-        reorg_mon = GraphQLReorganisationMonitor(graphql_url=urljoin(json_rpc_url, "/graphql"), check_depth=check_depth)
+        graphql_url = get_graphql_url(provider)
+        reorg_mon = GraphQLReorganisationMonitor(graphql_url=graphql_url, check_depth=check_depth)
     else:
         # Default slow implementation
         logger.warning("The node does not support /graphql interface. " "Downloading block headers and timestamps will be extremely slow." "Check documentation how to configure your node or choose a smaller timeframe for the buffer of trades.")

eth_defi/provider/ankr.py

@@ -1,8 +1,13 @@
 """Ankr specific Web3.py functionality.
 
 - Ankr has issues with some JSON-RPC access patterns.
-  See also :py:mod:`eth_defi.provider.broken_provider`.
 
+.. warning ::
+
+    We do not recommend using Ankr as it randomly returns empty responses to eth_call RPC method
+    and this is not fixable at the client side.
+
+See also :py:mod:`eth_defi.provider.broken_provider`.
 """
 from web3 import Web3
 

eth_defi/provider/llamarpc.py

@@ -0,0 +1,23 @@
+"""Fixes for quirks and features on LlamaNodes.
+
+- `LlamaNodes <https://llamanodes.com/>`__ runs RPC services at ``llamarpc.com``
+
+- Their RPC nodes have some compatibility issues we address in this module
+
+See also :py:mod:`eth_defi.provider.broken_provider`.
+"""
+from requests import Response
+
+
+def is_llama_bad_grapql_reply(resp: Response):
+    """Is the web server response fake 404 response from llamarpc.com
+
+    llamarpc.com web server does not know how to use HTTP 404 status code.
+
+    See :py:func:`eth_defi.chain.has_graphql_support`.
+    """
+    try:
+        content = resp.json()
+        return content.get("error").get("message") == 'UserKey was not a ULID or UUID'
+    except Exception:
+        return False

tests/provider/test_llamarpc.py

@@ -0,0 +1,19 @@
+"""LlamaNodes specific tests"""
+
+import os
+
+import pytest
+from web3 import Web3, HTTPProvider
+
+from eth_defi.chain import has_graphql_support
+
+JSON_RPC_LLAMA = os.environ.get("JSON_RPC_LLAMA")
+pytestmark = pytest.mark.skipif(not JSON_RPC_LLAMA, reason="This test needs LlamaNodes come node via JSON_RPC_LLAMA")
+
+
+def test_llama_is_bad():
+    """Work around fake 404 response from llamarpc.com"""
+    provider = HTTPProvider(JSON_RPC_LLAMA)
+    web3 = Web3(provider)
+    assert not has_graphql_support(provider)
+    assert web3.eth.block_number > 1