# Elastic search for time series

**URL:** <https://discuss.elastic.co/t/elastic-search-for-time-series/198600>\
**Category:** Elasticsearch\
**Created:** [September 9, 2019, 5:15am UTC](https://discuss.elastic.co/t/elastic-search-for-time-series/198600 "2019-09-09T05:15:36Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Mansoor\_Ali](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/mansoor_ali/32/53676_2.png) [@Mansoor\_Ali](https://discuss.elastic.co/u/Mansoor_Ali)\
**Post date:** [September 9, 2019, 5:15am UTC](https://discuss.elastic.co/t/elastic-search-for-time-series/198600/1 "2019-09-09T05:15:36Z")

</div>

Hi All,

I am new to Nosql and Elastic Search world.

I am evaluating Database for keeping 5 years of time series data for multiple clients, the data changes from client to client that’s why I am looking for NoSql Database like Elastic Search.

The data is mostly float numbers and we will be crushing numbers quite a lot to get the reporting. We keep data day by day and We need to crunch data of 2 years in a single go.

We have tried using CosmosDB from Microsoft and there are limitations on how much we can do with it and there is no proper aggregation in the Cosmos. I have read about Elastic Search would love to know in detail how the sql query does aggregations perform ?  
Is the index performance if we add a new column in the document for a year or two, does it affect the performance while the index is being created in the background ?

I also read about approximations that’s done across shard, the reports generated will require exact values, not approximates, does it causes a significant difference if I turn off approximation (somehow) ?

Thanks in Advance

---

<div class="post-metadata">

**Author:** ![Christian\_Dahlqvist](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/christian_dahlqvist/32/4617_2.png) [@Christian\_Dahlqvist](https://discuss.elastic.co/u/Christian_Dahlqvist)\
**Post date:** [September 9, 2019, 5:54am UTC](https://discuss.elastic.co/t/elastic-search-for-time-series/198600/2 "2019-09-09T05:54:39Z")

</div>

That is a very broad question, which will be difficult to answer.

Have you looked at the following resources?

> **[Elasticsearch as a Time Series Data Store](https://www.elastic.co/blog/elasticsearch-as-a-time-series-data-store)**
>
> Ever wondered how Elasticsearch handles time series metrics? Felix Barnsteiner from stagemonitor - an open source solution to application performance monitoring

> **[Querying and aggregating time series data in Elasticsearch](https://www.elastic.co/blog/querying-and-aggregating-time-series-data-in-elasticsearch)**
>
> Learn how to perform queries against time series data in Elasticsearch, including tips and tricks on aggregations and groupings.

> **[How to Create, Manage, and Visualize Elasticsearch Rollup Data in Kibana](https://www.elastic.co/blog/how-to-create-manage-and-visualize-elasticsearch-rollup-data-in-kibana)**
>
> Learn how to save space by rolling up historical Elasticsearch data into summary documents that you can create, manage, and visualize with new tools in Kibana.

---

<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:** [October 7, 2019, 5:54am UTC](https://discuss.elastic.co/t/elastic-search-for-time-series/198600/3 "2019-10-07T05:54:44Z")

</div>

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