# Starting with snapshots when we already have a large amount of data?

**URL:** <https://discuss.elastic.co/t/starting-with-snapshots-when-we-already-have-a-large-amount-of-data/305781>\
**Category:** Elasticsearch\
**Tags:** slm-snapshot-lifecycle-management, snapshot-and-restore\
**Created:** [May 27, 2022, 8:52am UTC](https://discuss.elastic.co/t/starting-with-snapshots-when-we-already-have-a-large-amount-of-data/305781 "2022-05-27T08:52:58Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Andreas\_Falk](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/andreas_falk/32/82595_2.png) [@Andreas\_Falk](https://discuss.elastic.co/u/Andreas_Falk)\
**Post date:** [May 27, 2022, 8:52am UTC](https://discuss.elastic.co/t/starting-with-snapshots-when-we-already-have-a-large-amount-of-data/305781/1 "2022-05-27T08:52:58Z")

</div>

Hi,

We shall do an upgrade from 7.10.2 to 7.17+ and have tried this is lab env.  
But one question that comes to mind is, how is the cluster going to react when starting a snapshot on over 10TB data with cold storage?

We are using an NFS share as snapshot storage, and the cold data are almost 8TB of the total data.

Is this going to make a huge performance footprint on the cluster as a whole, or is it just going to take a huge amount of time 🙂

HOT/WARM: 1.4TB  
COLD: 7.5TB

 ![image](https://us1.discourse-cdn.com/elastic/original/3X/f/a/fa43d137a0a484c008921769e4623cdfc71280ac.png)

One thought was to exclude the COLD data for snapshots, but it seems that I cant find if it is possible to make such exclusions.

--  
Kind Regards Falk

---

<div class="post-metadata">

**Author:** ![DavidTurner](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/davidturner/32/22453_2.png) [@DavidTurner](https://discuss.elastic.co/u/DavidTurner)\
**Post date:** [May 27, 2022, 1:11pm UTC](https://discuss.elastic.co/t/starting-with-snapshots-when-we-already-have-a-large-amount-of-data/305781/2 "2022-05-27T13:11:09Z")

</div>

> [@Andreas\_Falk](#):
>
> Is this going to make a huge performance footprint on the cluster as a whole, or is it just going to take a huge amount of time 🙂

The default snapshot throttle is 40MBps per node which should limit the performance impact. It'll only take a couple of days to create a snapshot of that size at 40MBps. Obviously you can adjust the throttle too, with faster speeds taking less time but having more performance impact.

---

<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:** [June 24, 2022, 1:11pm UTC](https://discuss.elastic.co/t/starting-with-snapshots-when-we-already-have-a-large-amount-of-data/305781/3 "2022-06-24T13:11:19Z")

</div>

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