Thanks, Ibo. That must be quite a lot of ids for the "live" query (flywire_partner_summary) to take ~4m and the whole CAVE+itanna server infrastructure is not completely robust when you start sending a lot of queries. A few options
1. try adding the retry argument eg retry=3. This doesn't apply to all of the server endpoints but can help with some
2. try method="cave" (if you haven't already). This moves all of the query to CAVE rather than itanna in case those are the flaky servers
3. try the hidden fafbseg:::flywire_partners_cave() method with multiple ids at once (by default it only sends one neuron at a time because of possible row limits in the response, but this is very conservative).
or if you can with you application,
and you can also run
fafbseg:::flywire_connectome_data_message()
to be told where to put the resultant file. All the best,