fix(manager_persistent_snapshots): increase expected timeouts

since manager 3.2 we have a "longer" restore operation,
since a repair is called as part of its operation (and
not manually by the user aftwards), making the `restore`
operation much longer it was before.

it should give time to the restore, and from staging
test on the smallest snapshot size, it took ~23 minutes
(instead of the 15 initially planned), and hence here
giving some extra, to accomodate different running
workloads.

defaults/manager_persistent_snapshots.yaml

@@ -4,7 +4,7 @@ aws:
   snapshots_sizes:
     5:
       number_of_rows: 5242880
-      expected_timeout: 900  # 15 minutes
+      expected_timeout: 1800  # 30 minutes
       snapshots:
         'sm_20230702185929UTC':
           keyspace_name: "5gb_sizetiered_2022_2"
@@ -23,7 +23,7 @@ aws:
           number_of_nodes: 5
     10:
       number_of_rows: 10485760
-      expected_timeout: 1800  # 30 minutes
+      expected_timeout: 3600  # 60 minutes
       snapshots:
         'sm_20230223105105UTC':
           keyspace_name: "10gb_sizetiered"
@@ -47,7 +47,7 @@ aws:
           number_of_nodes: 4
     100:
       number_of_rows: 104857600
-      expected_timeout: 9000  # 150 minutes
+      expected_timeout: 18000  # 300 minutes
       snapshots:
         'sm_20230223130733UTC':
           keyspace_name: "100gb_sizetiered"
@@ -71,7 +71,7 @@ aws:
           number_of_nodes: 4
     2048:
       number_of_rows: 2147483648
-      expected_timeout: 66000  # 1100 minutes
+      expected_timeout: 132000  # 2200 minutes
       snapshots:
         'sm_20230226074656UTC':
           keyspace_name: "2tb_sizetiered"