t/fetch_test: wait longer for leadership after stepping down

Drift in seastar lowres clock may lead to situation in which a leader
election will not trigger in 1.5 seconds. Increased a wait for leader
timeout to eliminate test spurious failures.

Fixes: redpanda-data#17943
Fixes: redpanda-data#18059

Signed-off-by: Michał Maślanka <michal@redpanda.com>

src/v/kafka/server/tests/fetch_test.cc

@@ -413,7 +413,7 @@ FIXTURE_TEST(fetch_leader_epoch, redpanda_thread_fixture) {
                   .get0();
             }
             partition->raft()->step_down("trigger epoch change").get0();
-            wait_for_leader(ntp).get0();
+            wait_for_leader(ntp, 10s).get0();
             {
                 auto batches = model::test::make_random_batches(
                   model::offset(0), 5);