fix(e2e): only update protocol parameters on leader aggregator

Since now the follower read the network config from the leader, this
means that the update of the protocol parameters is now a responsability
of the leader only.

This lead to flakiness because this step was restarting all aggregators,
and sometimes the follower started before the leader and had a error
when it executed its handle discrepencies because the leader http server
was still down.

Author: Alenar

mithril-test-lab/mithril-end-to-end/src/end_to_end_spec.rs

@@ -136,7 +136,9 @@ impl Spec {
             "epoch after which the protocol parameters will change".to_string(),
         )
         .await?;
-        assertions::update_protocol_parameters(aggregator).await?;
+        if aggregator.is_first() {
+            assertions::update_protocol_parameters(aggregator).await?;
+        }
 
         // Wait 6 epochs after protocol parameters update, so that we make sure that we use new protocol parameters as well as new stake distribution a few times
         target_epoch += 6;