Elasticsearch query failing with downsampling index

Elastic Stack Elasticsearch
1

We are using ELK 7.9.0 with downsampling turned on at 7 days. I have this query that works when the range is before the downsampling ocured. But if its in that range it gives me this exception.

This is the query

curl -u sysadmin:sysadmin -k -i https://<lab>:31001/mon-elasticsearch-client/metric*/_search?pretty -H 'Content-Type: application/json' -d '
{
 "query": {
    "bool": {
      "filter": [
        {
          "range": {
            "@timestamp": {
              "gte": "now-10d",
              "lte": "now-9d"
            }
          }
        },
        {
          "term": {
            "metricset.name": "uptime"
          }
        },
        {
          "term": {
            "system.uid": "aa757b27-1234-1234-1234-bbb4380e05fd"
          }
        }
      ]
    }
  },
  "aggregations": {
    "hosts": {
      "terms": {
        "field": "host.name",
        "size": 10000
      },
      "aggs": {
        "uptimes_agg": {
          "date_histogram": {
            "field": "@timestamp",
            "fixed_interval": "10m",
            "extended_bounds": {
              "min": "now-10d",
              "max": "now-9d"
            }
          },
          "aggs": {
            "max_uptime": {
              "max": {
                "field": "system.uptime.duration.ms",
                "missing": 0
              }
            }
          }
        }
      }
    }
  },
  "size": 0,
  "track_total_hits": false
}'

And this is the error I get

date: Wed, 26 Jun 2024 20:29:46 GMT
content-type: application/json
x-elastic-product: Elasticsearch
strict-transport-security: max-age=15724800; includeSubDomains

{
  "took" : 5,
  "timed_out" : false,
  "_shards" : {
    "total" : 36,
    "successful" : 32,
    "skipped" : 32,
    "failed" : 4,
    "failures" : [
      {
        "shard" : 0,
        "index" : "downsample-5m.ds-metricbeat-8.9.0-base-2024.06.14-000029",
        "node" : "z3jbbGoeQ6avEnFriA50yg",
        "reason" : {
          "type" : "illegal_argument_exception",
          "reason" : "Can't apply missing values on a class org.elasticsearch.xpack.aggregatemetric.aggregations.support.AggregateMetricsValuesSource$AggregateDoubleMetric$Fielddata"
        }
      },
      {
        "shard" : 0,
        "index" : "downsample-5m.ds-metricbeat-8.9.0-base-2024.06.16-000032",
        "node" : "_x0XtbSzQpqKLy7hcr-Qeg",
        "reason" : {
          "type" : "illegal_argument_exception",
          "reason" : "Can't apply missing values on a class org.elasticsearch.xpack.aggregatemetric.aggregations.support.AggregateMetricsValuesSource$AggregateDoubleMetric$Fielddata"
        }
      }
    ]
  },
  "hits" : {
    "total" : {
      "value" : 0,
      "relation" : "eq"
    },
    "max_score" : 0.0,
    "hits" : [ ]
  }
}

If I remove missing from the the max_uptime aggregate the query does not produced an error. Is there a way to get around this without loosing this setting of 0? Max aggregation | Elasticsearch Guide [8.14] | Elastic.