Almost 3 years later... I seem to be having the same issue you were having.
I have a worker pool, and each concurrent worker has its own client /
transport.
I do this because each transport/dialer has different proxy settings, and
there seemed to be no other way to circumvent this, such as
So far it looks like adding a timeout to the clients "works". Is this what
you ended up having to do?
On Monday, December 14, 2015 at 11:56:32 AM UTC-7, Sander van Harmelen
wrote:
>
> > While the http.Client itself isn't as big an issue, you will want to
> > pair each transport with its own