Recommended storage for using Discourse

AI-generated summary

Discourse users are discussing recommended storage solutions for hosting a forum. marianord suggests using fast storage, such as local storage for small installations, and S3 for larger setups with many attachments. Jonathan5 recommends using SSDs over spinning disks for improved performance, with marianord adding that NVMe is an even better option.

The original poster, romano97002, is looking for a turnkey hosting solution, specifically from Communiteq, and is seeking guidance on storage. Mevo clarifies that the discussion is about the type of storage, not the amount. codinghorror suggests asking the hosting service for guidance and provides a simple method for estimating storage needs based on user activity.

Good morning all,

I would like to know what storage is recommended to open its Discourse forum?

Thank you in advance for your reply.

Good afternoon.

That depends. The fastest the storage the better, usually for small installations I would go with the local storage, if you plan to have a lot of attachments it’s cheaper and easier to run in the long run to setup S3. That way the important stuff is fast enough and predictable and the media gets loaded via Object Storage.

Does this help? SSD rather than spinning disks: 502 error on nginx - #8 by Stephen

Yes, SSD helps a lot, NVMe is even better.

Hi,

Thank you for your answers.

I don’t know how to do it. It’s complex for me. I ask you this because I intend to choose the Discourse hosting solution which offers me turnkey creation and hosting.

Thank you in advance for your help.

Hi,

I bring the subject up to get an answer.

Thank you in advance for your help.

Didn’t you get an answer!?
Or was your question HOW MUCH storage, and not WHICH kind of storage? (I think “what” has been understood as “which” by most people, but maybe it wasn’t what you meant).

I suggest asking your hosting service?

You can also do some simple math – estimate how many pictures / attachments will be uploaded each day by how many users, then multiply by the number of days, etc.