Data server sizing-configuration guidelines

The inclusion of the IBM® Content Classification server on the data server changed the required memory and processors. Data server sizing-configuration guidelines are provided here.

To support the auto-classification function, you must increase the virtual machine settings for processors and memory on data servers to these new minimums:
  • Memory: 16 GB
  • vCPUs: 4

Required storage on the data server for the extracted text (snippet cache) is roughly 13% of the size of the uncompressed content, or 130 GB per 1 TB. The size can vary depending on the object type. The maximum amount of extracted text per item is configured to not exceed 2 MB by default.

A data server with 2 TB of configured storage can support approximately 10 TB of managed storage, assuming 50 million data objects with an average size of 200 KB. Approximately 300 GB is metadata storage with the remaining 1,700 GB supporting a combination of full text and extracted text.