ILM: Index wildcard pattern does not work for retry API when there are step errors

[ppf2]

6.6.0

When there are no step errors for the indices requested using the retry api, it will return a response indicating as such.

POST logstash-app1*/_ilm/retry

{
  "error": {
    "root_cause": [
      {
        "type": "illegal_argument_exception",
        "reason": "cannot retry an action for an index [logstash-app1-2019.02.15.19-000002] that has not encountered an error when running a Lifecycle Policy"
      }
    ],
    "type": "illegal_argument_exception",
    "reason": "cannot retry an action for an index [logstash-app1-2019.02.15.19-000002] that has not encountered an error when running a Lifecycle Policy"
  },
  "status": 400
}

When there are actually step errors for the indices that are covered as part of the wildcard pattern, the same retry request throws a Bad Request error as if it no longer knows how to process the wildcard pattern. (If I then issue the retry against individual indices logstash-app1-2019.02.15.19-000001, logstash-app1-2019.02.15.19-000002, etc..), the retry works. So this is specific to using wildcard pattern to tell the retry API to retry on multiple indices covered by the wildcard pattern.

POST logstash-app1*/_ilm/retry

{
  "error": {
    "root_cause": [
      {
        "type": "illegal_argument_exception",
        "reason": "index [logstash-app1] does not exist"
      }
    ],
    "type": "illegal_argument_exception",
    "reason": "index [logstash-app1] does not exist"
  },
  "status": 400
}

[elasticmachine]

Pinging @elastic/es-core-features