Thread Freezing Issue with ld.discovery.Chain After Failed Requests

Hi team,

We're encountering a critical issue when using ld.discovery.Chain to fetch futures option chain constituents. When making multiple sequential calls to ld.discovery.Chain(universe).constituents, if one call fails with an LDError, subsequent calls cause the Python instance or thread to freeze.

Issue Description:

Multiple calls to ld.discovery.Chain(universe).constituents work fine individually
However, if one call fails and throws an LDError, the next call in the sequence freezes
The freeze occurs specifically at line 95 in lseg.data.delivery._stream._stream

Code Example:

if __name__ == '__main__':

    import lseg.data as ld

    ld.open_session()


    def _fetch_option_chain(universes: list[str]) -> None:

        for univers in universes:

            print(f"starts fetch option chain for univers '{univers}'")

            try:

                ld.discovery.Chain(univers).constituents

            except ld.errors.LDError:

                print(f"error occurred for univers '{univers}'")

            print(f"finished fetch option chain for univers '{univers}'")


    _fetch_option_chain(["0#CLQ25+", "0#CLU25+"])  #  freezing when fetching 0#CLU25+ after 0#CLQ25+ failed (one thread occupied)

    _fetch_option_chain(["0#CLU25+"])  #  works


    ld.close_session()

Current Workaround: We've temporarily resolved this by closing and reopening the session after each LDError, but this doesn't address the underlying threading issue and impacts performance.

Environment:

Python version: 3.13
lseg.data version: 2.1.1

Could you please investigate this threading/connection state issue? It appears that failed requests are not properly cleaning up their resources, causing subsequent requests to hang.