NO-ENDPOINT: Updated OpenSearch Schema (#139)

Updated OpenSearch schema to correct data stream name for the new format
and added schema definition of error indexes. Without the error index
definition, OpenSearch defaults to using 10 shards for the error index
which is very wasteful, with 5 primary and 5 replicas.

schema/policies/rollover_strategy.json

@@ -31,8 +31,8 @@
       }
     ],
     "ism_template": {
-      "index_patterns": ["*_results*"],
-      "priority": 101
+      "index_patterns": ["*_transient*"],
+      "priority": 300
     }
   }
 }

schema/templates/error.json

@@ -0,0 +1,22 @@
+{
+  "index_patterns": [
+    "*_error"
+  ],
+  "template": {
+    "settings": {
+      "number_of_shards": 1,
+      "number_of_replicas": 0
+    },
+    "mappings": {
+      "dynamic": false,
+      "properties": {
+        "client_id": {
+          "type": "keyword"
+        },
+        "data": {
+          "type": "text"
+        }
+      }
+    }
+  }
+}