# Metricbeat vs Telegraf

**URL:** <https://discuss.elastic.co/t/metricbeat-vs-telegraf/93365>\
**Category:** Beats\
**Tags:** metricbeat\
**Created:** [July 17, 2017, 9:21am UTC](https://discuss.elastic.co/t/metricbeat-vs-telegraf/93365 "2017-07-17T09:21:50Z")\
**Posts on this page:** 8\
**Page:** 1

<div class="post-metadata">

**Author:** ![mahadevaprasadap](https://avatars.discourse-cdn.com/v4/letter/m/59ef9b/32.png) [@mahadevaprasadap](https://discuss.elastic.co/u/mahadevaprasadap)\
**Post date:** [July 17, 2017, 9:21am UTC](https://discuss.elastic.co/t/metricbeat-vs-telegraf/93365/1 "2017-07-17T09:21:51Z")

</div>

I have a requirement to collect system and application metrics from servers. But before that any one please let me know Metricbeat vs Telegraf to understand better.

---

<div class="post-metadata">

**Author:** ![tudor](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/tudor/32/3753_2.png) [@tudor](https://discuss.elastic.co/u/tudor)\
**Post date:** [July 17, 2017, 10:21am UTC](https://discuss.elastic.co/t/metricbeat-vs-telegraf/93365/2 "2017-07-17T10:21:20Z")

</div>

Metricbeat is designed for Elasticsearch, while Telegraf is mainly used with Influxdb. The decision between the two is more of a decision between the Elastic stack and the TICK stack.

If you have the requirment, or might have in the future, to also collect logs and tracing data besides metrics, then the Elastic stack and Metricbeat are IMO a better choice, because the TICK stack is focused only on metrics.

---

<div class="post-metadata">

**Author:** ![mahadevaprasadap](https://avatars.discourse-cdn.com/v4/letter/m/59ef9b/32.png) [@mahadevaprasadap](https://discuss.elastic.co/u/mahadevaprasadap)\
**Post date:** [July 17, 2017, 10:52am UTC](https://discuss.elastic.co/t/metricbeat-vs-telegraf/93365/3 "2017-07-17T10:52:47Z")

</div>

Thank you for the quick response. If I need to create custom metrics (eg., need to collect the metrics of applications running in the server) , may I know which agent will be good?

---

<div class="post-metadata">

**Author:** ![TimWard](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/timward/32/19574_2.png) [@TimWard](https://discuss.elastic.co/u/TimWard)\
**Post date:** [July 17, 2017, 11:45am UTC](https://discuss.elastic.co/t/metricbeat-vs-telegraf/93365/4 "2017-07-17T11:45:08Z")

</div>

Depending on what the application is and what you want to measure there will be a number of choices.

One approach that will always work regardless of the above is to get the application to write out the metrics to a log file, and then use Filebeat on the application server to pick up the log files then Logstash and Elasticsearch on their own kit to process and store the data.

---

<div class="post-metadata">

**Author:** ![mahadevaprasadap](https://avatars.discourse-cdn.com/v4/letter/m/59ef9b/32.png) [@mahadevaprasadap](https://discuss.elastic.co/u/mahadevaprasadap)\
**Post date:** [July 17, 2017, 1:22pm UTC](https://discuss.elastic.co/t/metricbeat-vs-telegraf/93365/5 "2017-07-17T13:22:56Z")

</div>

Thank you Tim.

I want to monitor the application's data base(PostgreSQL). For example, for number of users logged in from PostgreSQL table

---

<div class="post-metadata">

**Author:** ![mahadevaprasadap](https://avatars.discourse-cdn.com/v4/letter/m/59ef9b/32.png) [@mahadevaprasadap](https://discuss.elastic.co/u/mahadevaprasadap)\
**Post date:** [July 18, 2017, 4:59am UTC](https://discuss.elastic.co/t/metricbeat-vs-telegraf/93365/6 "2017-07-18T04:59:21Z")

</div>

Also expecting differences in terms of Performance/Reliability , Community Support, Ease of use etc.,

---

<div class="post-metadata">

**Author:** ![rpsing](https://avatars.discourse-cdn.com/v4/letter/r/96bed5/32.png) [@rpsing](https://discuss.elastic.co/u/rpsing)\
**Post date:** [July 20, 2017, 9:20pm UTC](https://discuss.elastic.co/t/metricbeat-vs-telegraf/93365/7 "2017-07-20T21:20:32Z")

</div>

I looked at the input plugins supported by Telegraf, and it seems like its a much more comprehensive list than metricbeat. For eg, it includes support for MSSQL, and SNMP polling. Given that Telegraf can output in JSON and has a flexible output plugin architecture, any reason why metricbeat might still be a better option?

---

<div class="post-metadata">

**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\
**Post date:** [August 17, 2017, 9:20pm UTC](https://discuss.elastic.co/t/metricbeat-vs-telegraf/93365/8 "2017-08-17T21:20:44Z")

</div>

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.
