Can I update my gRPC middleware in a ‘live’ scenario? That is, if I take my middleware offline (break the gRPC socket) and restart it, will requests queue up while the middleware is down - then fail after a timeout? Or will all requests fail while the middleware is unavailable.
Hi, we’re using the default gRPC client settings, in case of failure Tyk will keep reconnecting to your server, meanwhile new requests will result in a middleware error.
We may adjust this functionality in the future and allow the usage of retry policies.
Thanks