# Slow performance compare to v1.7 when using 5.x

**URL:** https://discuss.elastic.co/t/slow-performance-compare-to-v1-7-when-using-5-x/101506
**Category:** Elasticsearch
**Created:** [September 22, 2017, 2:33pm UTC](https://discuss.elastic.co/t/slow-performance-compare-to-v1-7-when-using-5-x/101506 "2017-09-22T14:33:28Z")
**Posts on this page:** 4
**Page:** 1

<div class="post-metadata">

### Author: ![itplayer](https://avatars.discourse-cdn.com/v4/letter/i/d26b3c/32.png) [@itplayer](https://discuss.elastic.co/u/itplayer)
#### Post date: [September 22, 2017, 2:33pm UTC](https://discuss.elastic.co/t/slow-performance-compare-to-v1-7-when-using-5-x/101506/1 "2017-09-22T14:33:28Z")

</div>

I found an issue that might be the root cause

[Slow performance on HDD #16676](https://github.com/elastic/elasticsearch/issues/16676)

Can anyone help to explain what is the reason behind the `index.translog.durability`  
was added since v2.x?

Does v1.7 default behavior equals to `index.translog.durability: async`

What is the benefit to set `index.translog.durability: request`?

Looking forward your input. Thanks

---

<div class="post-metadata">

### Author: ![dadoonet](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/dadoonet/32/137187_2.png) [@dadoonet](https://discuss.elastic.co/u/dadoonet)
#### Post date: [September 22, 2017, 2:56pm UTC](https://discuss.elastic.co/t/slow-performance-compare-to-v1-7-when-using-5-x/101506/2 "2017-09-22T14:56:02Z")

</div>

`index.translog.durability: async` is what you had in 1.7 indeed.

Setting that can cause data loss. So activate that only if you don't care about your data.

Then use bulk API and the price of this change should be minimal IMO.  
Use SSD drives if you are not using them yet.

---

<div class="post-metadata">

### Author: ![itplayer](https://avatars.discourse-cdn.com/v4/letter/i/d26b3c/32.png) [@itplayer](https://discuss.elastic.co/u/itplayer)
#### Post date: [September 22, 2017, 3:08pm UTC](https://discuss.elastic.co/t/slow-performance-compare-to-v1-7-when-using-5-x/101506/3 "2017-09-22T15:08:02Z")

</div>

Thanks for your reply. Since we have `async` in v1.7 and we use 1.7 for a while without critical data problem, I think I can take that price.

Of course, to use bulk operation would be next step to enhance. Finally, I got this setting, at first I thought that was merge policy settings problem cause the problem.

---

<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 20, 2017, 3:08pm UTC](https://discuss.elastic.co/t/slow-performance-compare-to-v1-7-when-using-5-x/101506/4 "2017-10-20T15:08:53Z")

</div>

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