[URL]
https://github.com/openshift/openshift-docs/blob/enterprise-4.7/modules/cluster-logging-deploy-storage-considerations.adoc

[ISSUE]
In the documentation, it's possible to read ". "Some Elasticsearch users have found that it is necessary to keep absolute storage consumption around 50% and below 70% at all times."

This is not a good recommendation really. It depends absolutely on the size of the storage:

If you have below 70% used. The 70% for:

• 10G is 3G free
• 100G is 30G free
• 1TB is 300G free
• 2TB is 600G free

If they are around 50% of use. The 50% for:

• 10G is 5G
• 100G is 50G
• 1TB is 500G
• 2TB is 1TB

Then, that recommendation is not a good recommendation. It depends absolutely of the size of the storage that each user has.

I'll recommend leaving out of the documentation that phrase.

On the other hand, I don't know exactly how the previous is helping with this:

"This helps to avoid Elasticsearch becoming unresponsive during large merge operations."

If it's doing large merge operations, it's known that it's going to need to create some temporary queries, but, I believe that the most problematic to get "unresponsive" during large merge operations is how large are your data and how much is the performance of your infra.

Then, at the same time, I'd leave out of the documentation that phrase or it could be expressed in a different way, but not related to the percentage of the storage.

This was reported for OCP 4.6 in BZ#1905583

Best regards,
Oscar