# Metricbeat 7.17.20 cannot index logstash metrics into Elasticsearch
**URL:** https://discuss.elastic.co/t/metricbeat-7-17-20-cannot-index-logstash-metrics-into-elasticsearch/358275
**Category:** Beats
**Tags:** docker, metricbeat
**Created:** [April 26, 2024, 8:12am UTC](https://discuss.elastic.co/t/metricbeat-7-17-20-cannot-index-logstash-metrics-into-elasticsearch/358275 "2024-04-26T08:12:41Z")
**Posts on this page:** 2
**Page:** 1
### Author:  [@gergelyzsamboki-seon](https://discuss.elastic.co/u/gergelyzsamboki-seon)
#### Post date: [April 26, 2024, 8:12am UTC](https://discuss.elastic.co/t/metricbeat-7-17-20-cannot-index-logstash-metrics-into-elasticsearch/358275/1 "2024-04-26T08:12:41Z")
i'm using metricbeat to monitor logstash (also 7.17.20). both are running in docker.
Metricbeat is logging the following error:
`mapper [logstash_stats.pipelines.events.duration_in_millis] cannot be changed from type [float] to [long]`
Metricbeat configuration:
```auto
metricbeat.config.modules:
path: ${path.config}/modules.d/*.yml
reload.enabled: false
processors:
- add_cloud_metadata: ~
- add_docker_metadata: ~
output.elasticsearch:
hosts: https://es-log.domain.internal:9200
username: logstash-log
password: ...
name: ls-dev-log-0-i-0182be70eacab9ae8.euw1
monitoring:
cluster_uuid: ...
enabled: true
logging.level: debug
```
Module config:
```auto
- module: logstash
enabled: true
xpack.enabled: true
period: 10s
hosts: ["http://logstash:9600"]
```
The data:
```auto
2024-04-25T15:28:56.979Z DEBUG [processors] processing/processors.go:203 Publish event: {
"@timestamp": "2024-04-25T15:28:56.912Z",
"@metadata": {
"beat": "metricbeat",
"type": "_doc",
"version": "7.17.20",
"index": ".monitoring-logstash-7-mb"
},
"interval_ms": 10000,
"logstash_stats": {
"events": {
"duration_in_millis": 4.7338102e+07,
"in": 4.4495605e+07,
"filtered": 4.4495605e+07,
"out": 4.3300184e+07
},
```
why are the numbers in floating point format? is there a way to turn that off?
---
### Author:  [@gergelyzsamboki-seon](https://discuss.elastic.co/u/gergelyzsamboki-seon)
#### Post date: [April 26, 2024, 2:51pm UTC](https://discuss.elastic.co/t/metricbeat-7-17-20-cannot-index-logstash-metrics-into-elasticsearch/358275/2 "2024-04-26T14:51:07Z")
it turned out the issue was in the elasticsearch template. we have a default index template which unfortunately overlaped the default monitoring templates.
`index template [default-logs] has index patterns [*logs*] matching patterns from existing older templates [.monitoring-logstash,.monitoring-es,logstash,ecs-logstash,.monitoring-beats,.items-default,.lists-default,.monitoring-kibana] with patterns (.monitoring-logstash => [.monitoring-logstash-7-*],.monitoring-es => [.monitoring-es-7-*],logstash => [logstash-*],ecs-logstash => [ecs-logstash-*],.monitoring-beats => [.monitoring-beats-7-*],.items-default => [.items-default-*],.lists-default => [.lists-default-*],.monitoring-kibana => [.monitoring-kibana-7-*]); this template [default-logs] will take precedence during new index creation`
changing the pattern on our template fixed the issue