Error 429: catch Retry-After

[Hadevmin]

I would like to know how exploit the return of the server when an error 429 occurs.

I use this template to perform my gql requests:

        async with Client(transport=transport) as session:
            query = gql(query)
            params = {params}
            try:
                result = await session.execute(query, variable_values=params)
            except TransportServerError:
                time.sleep(x_seconds)

I would like to set the x_seconds with the return of the server.

Whenever you perform too many requests, the server will answer with a 429 HTTP error code and add a 'Retry-After: TimeToWaitInSeconds' header to the response. (RFC: https://datatracker.ietf.org/doc/html/rfc6585#section-4)

Should I use the requests library to catch this parameter or can I use the gql library to catch it?

Thanks

[leszekhanusz]

You should have the response headers in the transport.response_headers attribute, so you could get the value you're looking for in transport.response_headers["Retry-After"]

This is already working with the RequestsHTTPTransport but for the AIOHTTPTransport the response headers were only saved if there was no error.

The PR #367 should fix this for this transport as well.

[Hadevmin]

Thanks you for your help!