Hi there, I have a quick question: When enabling

--enable-search-analytics

, an additional

analytics-dir

volume must be mounted on each node. What should be considered regarding the size of this volume?

As of today we don't actually use this for anything more than for some metadata keeping because analytics data is primarily aggregated onto other collections. However, in future we are going to add an option to log the queries etc. and that's where this will make a difference.

Thanks for the response. It is mentioned though in the documentation, that you enable it automatically for all clusters in Typesense Cloud and I would like to do so in Kubernetes(via the Typesense Kubernetes Operator). For that matter I would like to understand a bit better what size should be the size of the analytics PVC and if it relates to the size of the data PVC or the memory allocated in any extent. Additionally, I would like to clarify if that volume could be a) ephemeral and b) if it is node-specific or shared cluster-wide. Thanks in advance.

The actual size will depend on how much logs the user might store. Maybe choose something small.