Default httpx timeout too short

[alyssadai]

Our current default timeout limit of 30s is now too short of a window for the n-API to receive the response data from graph, now that we have upgraded to a session-level graph data model (meaning the # triples has greatly increased).

Until we do some internal optimization, we should disable this timeout to avoid n-APIs always timing out for large graph stores.

api/app/api/crud.py

Lines 22 to 50 in 1f89c22

	def post_query_to_graph(query: str, timeout: float = 30.0) -> dict:
	"""
	Makes a post request to the graph API to perform a query, using parameters from the environment.
	# TODO: Revisit default timeout value when query performance is improved
	Parameters
	----------
	query : str
	The full SPARQL query string.
	timeout : float, optional
	The maximum duration for the request, by default 5.0 seconds.
	Returns
	-------
	dict
	The response from the graph API, encoded as json.
	"""
	try:
	response = httpx.post(
	url=util.QUERY_URL,
	content=query,
	headers=util.QUERY_HEADER,
	auth=httpx.BasicAuth(
	os.environ.get(util.GRAPH_USERNAME.name),
	os.environ.get(util.GRAPH_PASSWORD.name),
	),
	timeout=timeout,
	)

[surchs]

🚀 Issue was released in v0.2.1 🚀