roachtest: increase max accepted latency on c2c shutdown tests

The node shutdown tests occasionally fail when the latency exceeds 2 minutes;
however if the coordinator shutsdown, a new node may take up to two minutes to
adopt the job. This patch bumps the latency to 4 mintues to account for the job
system polling.

Informs #132730

Epic: none

pkg/cmd/roachtest/tests/cluster_to_cluster.go

@@ -1590,8 +1590,12 @@ func registerClusterReplicationResilience(r registry.Registry) {
 			cutover:                              3 * time.Minute,
 			expectedNodeDeaths:                   1,
 			sometimesTestFingerprintMismatchCode: true,
-			clouds:                               registry.OnlyGCE,
-			suites:                               registry.Suites(registry.Nightly),
+			// The job system can take up to 2 minutes to reclaim a job if the
+			// coordinator dies, so increase the max expected latency to account for
+			// our lovely job system.
+			maxAcceptedLatency: 4 * time.Minute,
+			clouds:             registry.OnlyGCE,
+			suites:             registry.Suites(registry.Nightly),
 		}
 
 		c2cRegisterWrapper(r, rsp.replicationSpec,