Elasticsearch explicit field mappings in template not taken when data is logged from filebeat through ingest pipeline

[shi]

sir,
I have created an index life cycle policy with name

lp_ilp_fgnew1

Created index template lp_templateforotherfgutmwaf7 as follows

PUT _index_template/lp_templateforotherfgutmwaf7
{
  "index_patterns": ["lp_index-otherfgutmwaf7-*"],                 
  "template": {
    "settings": {
      "number_of_shards": 1,
      "number_of_replicas": 1,
      "index.lifecycle.name": "lp_ilp_fgnew1",      
      "index.lifecycle.rollover_alias": "lp_index-otherfgutmwaf7"    
    }
  }
}  

Bootstrapped initial indices for lp_index-otherfgutmwaf as follows
I have tried to explicitly map some of the fields also in the mappings section here

PUT /%3Clp_index-otherfgutmwaf7-%7Bnow%2Fd%7D-000001%3E
{
"aliases": {
    "lp_index-otherfgutmwaf7": {
      "is_write_index": true
    }
  },
  "mappings" : {
      "properties" : {
        "action" : {
              "type" : "keyword"
        },
        "direction" : {
          "type" : "text"
        },
        "dstintf" : {
          "type" : "integer"
        }
      },
      "dynamic": true
    }
}

Created ingest pipe line lp_index-otherfgutmwaf7-pipeline as follows

PUT _ingest/pipeline/lp_index-otherfgutmwaf7-pipeline
{
  "description": "Ingest pipeline",
  "processors": [
    {
      "dissect": {
        "field": "message",
        "pattern" : "%{timestamp} _gateway %{kvmsg}"
      }
    },
    {
      "kv": {
    "field": "kvmsg",
    "field_split": " (?=[a-z\\_\\-]+=)",
    "value_split": "=",
    "trim_value": "\""
  }
    },
    {
      "date": {
        "field": "timestamp",
        "formats": [
          "ISO8601"
        ],
        "output_format": "yyyy-MM-dd'T'HH:mm:ss.SSSSSSSSSXXX"
      }
    },
    {
      "remove": {
        "field": "timestamp",
        "ignore_missing": true
      }
    },
    {
      "remove": {
        "field": "message",
        "ignore_missing": true
      }
    },  
    {
      "remove": {
        "field": "kvmsg",
        "ignore_missing": true
      }
    },
    {
      "remove": {
        "field": "devname",
        "ignore_missing": true
      }
    },
    {
      "remove": {
        "field": "devid",
        "ignore_missing": true
      }
    },
    {
      "remove": {
        "field": "type",
        "ignore_missing": true
      }
    }
  ]
}

Log data is populated from filebeat to the index through the pipeline.

Log data population to the index is also tested using direct post command as follows

{
  "action": "block",
  "direction": "output",
  "testfield": "Testing",
  "dstintf": "6"
}

Log data from file gets loaded in index from file beat through pipeline only when the explicit mappings section in index template is not there. . Using direct post command also works.

When explicit mappings section in index template is there even when file data DOES NOT gets loaded in index from file beat through pipeline , it gets loaded
using direct post command.

Why is it so ?
How to load Log data from file in index from file beat through pipeline when the explicit mappings section in index template is there?

thanks and regards
shini
How to

[stephenb (Stephen Brown)]

Please share your filebeat.yml

[shi]

Sir,
It is now understood that from filebeat though pipeline the issue was with

 "dstintf" : {
          "type" : "integer"
        }

It was my mistake ,
actual value of dstintf from log file is port6 and not 6 which was used while successfully testing using the POST command.
So filebeat was throwing integer parsing error or this field and it failed. I could see all these error messaages in filebeat clearly after I commented out the modules section in filebeat which was also generating lots of logs.

The issue is solved
Thanks and regards,
shini