[BUG] [alerting_exception] analyzer [analyzer_keyword] has not been configured in mappings

[paasi6666]

What is the bug?
When defining a new monitor (under alerting) and selecting the type 'Per document monitor', the monitor saves with following error:

Furthermore, when "testing" the query it times out:

The query looks like this:

and as shown below, the query works when discovering:

How can one reproduce the bug?
Steps to reproduce the behavior:

1. Go to Alerting>Monitors>Create monitor
2. Select 'Per document monitor', select any index and choose a query
3. Go to Preview query and performance and wait..
4. Try to save the monitor

What is the expected behavior?
I don't know, hence it never worked for me.

What is your host/environment?

• OS: Centos7
• Opensearch Version: 2.7.0
• Opensearch-Dashboards Version: 2.7.0

NOTE
We are ingesting the logs using graylog.

[AWSHurneyt]

Transferred this issue to the Alerting plugin folder as alerting owns development and maintenance of Document level monitors.

[lezzago]

@paasi6666 Could you please share your index mapping for us to reproduce the issue?

[paasi6666]

Sure. Note, that the mapping is generated by graylog.

{
  "rpz_0": {
    "mappings": {
      "dynamic_templates": [
        {
          "internal_fields": {
            "match": "gl2_*",
            "match_mapping_type": "string",
            "mapping": {
              "type": "keyword"
            }
          }
        },
        {
          "store_generic": {
            "match_mapping_type": "string",
            "mapping": {
              "type": "keyword"
            }
          }
        }
      ],
      "properties": {
        "@metadata_beat": {
          "type": "keyword"
        },
        "@metadata_type": {
          "type": "keyword"
        },
        "@metadata_version": {
          "type": "keyword"
        },
        "@timestamp": {
          "type": "date"
        },
        "agent_ephemeral_id": {
          "type": "keyword"
        },
        "agent_name": {
          "type": "keyword"
        },
        "beats_type": {
          "type": "keyword"
        },
        "client_id": {
          "type": "keyword"
        },
        "event_action": {
          "type": "keyword"
        },
        "full_message": {
          "type": "text",
          "analyzer": "standard"
        },
        "gl2_accounted_message_size": {
          "type": "long"
        },
        "gl2_message_id": {
          "type": "keyword"
        },
        "gl2_processing_error": {
          "type": "keyword"
        },
        "gl2_processing_timestamp": {
          "type": "date",
          "format": "uuuu-MM-dd HH:mm:ss.SSS"
        },
        "gl2_receive_timestamp": {
          "type": "date",
          "format": "uuuu-MM-dd HH:mm:ss.SSS"
        },
        "gl2_remote_ip": {
          "type": "keyword"
        },
        "gl2_remote_port": {
          "type": "long"
        },
        "gl2_source_input": {
          "type": "keyword"
        },
        "gl2_source_node": {
          "type": "keyword"
        },
        "host_name": {
          "type": "keyword"
        },
        "hostname": {
          "type": "keyword"
        },
        "log_file_path": {
          "type": "keyword"
        },
        "log_offset": {
          "type": "long"
        },
        "loglevel": {
          "type": "keyword"
        },
        "message": {
          "type": "text",
          "analyzer": "standard"
        },
        "query_action": {
          "type": "keyword"
        },
        "query_class": {
          "type": "keyword"
        },
        "query_name": {
          "type": "keyword"
        },
        "query_type": {
          "type": "keyword"
        },
        "rpz_category": {
          "type": "keyword"
        },
        "rpz_message": {
          "type": "keyword"
        },
        "rpz_zone": {
          "type": "keyword"
        },
        "source": {
          "type": "text",
          "analyzer": "analyzer_keyword",
          "fielddata": true
        },
        "source_ip": {
          "type": "keyword"
        },
        "source_port": {
          "type": "keyword"
        },
        "streams": {
          "type": "keyword"
        },
        "timestamp": {
          "type": "date",
          "format": "uuuu-MM-dd HH:mm:ss.SSS"
        },
        "url_domain": {
          "type": "keyword"
        },
        "url_short": {
          "type": "keyword"
        }
      }
    }
  }
}

[paasi6666]

@lezzago After taking a look at the index mapping, I see the issue:

"source": {
          "type": "text",
          "analyzer": "analyzer_keyword",
          "fielddata": true
        },

How do i update this to work as intended?

[paasi6666]

@lezzago The settings of the index look like follows:

GET rpz_0/_settings
{
  "rpz_0": {
    "settings": {
      "index": {
        "number_of_shards": "4",
        "provided_name": "rpz_0",
        "creation_date": "1649938793819",
        "analysis": {
          "analyzer": {
            "analyzer_keyword": {
              "filter": "lowercase",
              "tokenizer": "keyword"
            }
          }
        },
        "number_of_replicas": "0",
        "uuid": "e8NRlQCHQfau984C3QGMPQ",
        "version": {
          "created": "7100299",
          "upgraded": "136287827"
        }
      }
    }
  }
}

[pawelw1]

This issue has been reported in the OpenSearch forum https://forum.opensearch.org/t/alerting-exception-analyzer-analyzer-keyword-has-not-been-configured-in-mappings/14777.

I've tested this issue with an example from Elastic https://www.elastic.co/guide/en/elasticsearch/reference/current/analyzer.html

The result was exactly the same as reported in the forum case.

opensearch-node1_2.6.0   | org.opensearch.alerting.util.AlertingException: analyzer [my_analyzer] has not been configured in mappings
opensearch-node1_2.6.0   |      at org.opensearch.alerting.util.AlertingException$Companion.wrap(AlertingException.kt:70) ~[opensearch-alerting-2.6.0.0.jar:2.6.0.0]
opensearch-node1_2.6.0   |      at org.opensearch.alerting.util.DocLevelMonitorQueries.updateQueryIndexMappings(DocLevelMonitorQueries.kt:369) ~[opensearch-alerting-2.6.0.0.jar:2.6.0.0]
opensearch-node1_2.6.0   |      at org.opensearch.alerting.util.DocLevelMonitorQueries.access$updateQueryIndexMappings(DocLevelMonitorQueries.kt:45) ~[opensearch-alerting-2.6.0.0.jar:2.6.0.0]
opensearch-node1_2.6.0   |      at org.opensearch.alerting.util.DocLevelMonitorQueries$updateQueryIndexMappings$1.invokeSuspend(DocLevelMonitorQueries.kt) ~[opensearch-alerting-2.6.0.0.jar:2.6.0.0]
opensearch-node1_2.6.0   |      at kotlin.coroutines.jvm.internal.BaseContinuationImpl.resumeWith(ContinuationImpl.kt:33) [kotlin-stdlib-1.6.10.jar:1.6.10-release-923(1.6.10)]
opensearch-node1_2.6.0   |      at kotlinx.coroutines.DispatchedTask.run(Dispatched.kt:285) [kotlinx-coroutines-core-1.1.1.jar:?]
opensearch-node1_2.6.0   |      at kotlinx.coroutines.scheduling.CoroutineScheduler.runSafely(CoroutineScheduler.kt:594) [kotlinx-coroutines-core-1.1.1.jar:?]
opensearch-node1_2.6.0   |      at kotlinx.coroutines.scheduling.CoroutineScheduler.access$runSafely(CoroutineScheduler.kt:60) [kotlinx-coroutines-core-1.1.1.jar:?]
opensearch-node1_2.6.0   |      at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.run(CoroutineScheduler.kt:742) [kotlinx-coroutines-core-1.1.1.jar:?]
opensearch-node1_2.6.0   | Caused by: java.lang.Exception: java.lang.IllegalArgumentException: analyzer [my_analyzer] has not been configured in mappings
opensearch-node1_2.6.0   |      ... 9 more
opensearch-node1_2.6.0   | [2023-06-21T15:04:14,379][ERROR][o.o.a.u.AlertingException] [opensearch-node1] Alerting error: AlertingException[analyzer [my_analyzer] has not been configured in mappings]; nested: Exception[java.lang.IllegalArgumentException: analyzer [my_analyzer] has not been configured in mappings];

I've tested that analyzer with examples from the link and it works with no issues but for some reason Alerting plugin can't see the definition of that analyzer when creating a Monitor.
Looking at the logs, it looks like the Analyzer plugin trigger errors when the index is selected in the Monitor and before hitting the create button. I assume there is a validation running before the Monitor is created.

[petardz]

This is likely because Alerting plugin doesn't copy analyzer def from source index settings to queryIndex settings.

[paasi6666]

Missing feature or bug? Maybe a OpenSearch developer can take a look at this?

[eirsep]

Analyzer updates are static config changes to an index. they would require closing an index > apply analyzer setting change > re-open index.

Closing alerting query index is not possible as all monitors share query index and monitors are running in parallel.
We cannot support this in current architecture

[mvanderlee]

@eirsep So any index with a custom analyzer or normalizer simply can not have a monitor configured on it? And this is acceptable to the OpenSearch team?!
Since this is tied to security detectors as well, this should be a considered a major bug, not to be closed lightly!

opensearch-project/security-analytics#697

[dblock]

Looks like this was reopened, thanks for your feedback @mvanderlee.

@eirsep If you have ideas of what needs to change to support this feature, please add.

[Catch All Triage w/ 1, 2, 3]