# Blue/Green Deployment to 7.7

**URL:** <https://discuss.elastic.co/t/blue-green-deployment-to-7-7/232780>\
**Category:** Elasticsearch\
**Created:** [May 15, 2020, 9:27am UTC](https://discuss.elastic.co/t/blue-green-deployment-to-7-7/232780 "2020-05-15T09:27:09Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![spike83](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/spike83/32/47036_2.png) [@spike83](https://discuss.elastic.co/u/spike83)\
**Post date:** [May 15, 2020, 9:27am UTC](https://discuss.elastic.co/t/blue-green-deployment-to-7-7/232780/1 "2020-05-15T09:27:09Z")

</div>

Hi,

I'm running a 3 node cluster running 1 shard 2 replica for most indexes. Is it possible to carry out a Blue/Green deployment type upgrade to Elasticsearch whereby I shutdown a node and bring up a brand new node on 7.7 which then replicates the data from the existing 2? then repeat this step until all 3 nodes are upgraded?

If so will i need to carry out any of the pre-work around sync flush, disable shard allocation etc...

Thanks

---

<div class="post-metadata">

**Author:** ![warkolm](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/warkolm/32/39224_2.png) [@warkolm](https://discuss.elastic.co/u/warkolm)\
**Post date:** [May 15, 2020, 9:40am UTC](https://discuss.elastic.co/t/blue-green-deployment-to-7-7/232780/2 "2020-05-15T09:40:51Z")

</div>

Yes it is!

---

<div class="post-metadata">

**Author:** ![spike83](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/spike83/32/47036_2.png) [@spike83](https://discuss.elastic.co/u/spike83)\
**Post date:** [May 15, 2020, 9:43am UTC](https://discuss.elastic.co/t/blue-green-deployment-to-7-7/232780/3 "2020-05-15T09:43:27Z")

</div>

will i need to carry out any of the pre-work around sync flush, disable shard allocation or can I treat it like a failed node and just shutdown and bring up a new node?

---

<div class="post-metadata">

**Author:** ![warkolm](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/warkolm/32/39224_2.png) [@warkolm](https://discuss.elastic.co/u/warkolm)\
**Post date:** [May 15, 2020, 9:46am UTC](https://discuss.elastic.co/t/blue-green-deployment-to-7-7/232780/4 "2020-05-15T09:46:11Z")

</div>

The cleanest way is to add the new node, then use allocation filtering to exclude one of the old version nodes, wait for it to be empty, then remove. Repeat.

Given you have 2 replicas, you can probably just "hard" stop one node and the cluster will be ok. Then add the new one and let it rebalance. But the first method is ultimately safer.

---

<div class="post-metadata">

**Author:** ![spike83](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/spike83/32/47036_2.png) [@spike83](https://discuss.elastic.co/u/spike83)\
**Post date:** [May 15, 2020, 10:25am UTC](https://discuss.elastic.co/t/blue-green-deployment-to-7-7/232780/5 "2020-05-15T10:25:56Z")

</div>

Great thanks

---

<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 12, 2020, 10:26am UTC](https://discuss.elastic.co/t/blue-green-deployment-to-7-7/232780/6 "2020-06-12T10:26:02Z")

</div>

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