all: retry the request more agressively

For better recovery and dealing with network stack changes, this commit
change the request flow to:

failure of any kind -> recreate transport/re-bootstrap -> retry once

That would make ctrld recover from all scenarios in theory.

resolver.go

@@ -93,8 +93,5 @@ func (r *legacyResolver) Resolve(ctx context.Context, msg *dns.Msg) (*dns.Msg, e
 		Dialer: dialer,
 	}
 	answer, _, err := dnsClient.ExchangeContext(ctx, msg, r.endpoint)
-	if _, ok := err.(*net.OpError); ok {
-		return answer, ErrUpstreamFailed
-	}
 	return answer, err
 }