Scaling cim pod CPU and memory using commits

The following sections contain tables that enable you to scale your CPU and memory resources to provide estimated commit throughput for various IDIR sizes and commit pool threads submitted. The estimated response time is called the commit throughput. The tables are a guide to scale your cloud hardware resources for expected performance given your commit count and IDIR size.

Important: In a deployment with multiple Coverity Connect (cim) pods, the resource guidance applies equally to all pods deployed. Provide the full resources for each pod that you deploy; do not split up calculated CPU and memory resources between pods.

To use the hardware scaling tables for commit performance:

Find your target intermediate directory (IDIR) size. Find either the largest IDIR or the slowest commit, and use the corresponding IDIR size on disk. Refer to the corresponding table:
- Medium – up to 2.6 GB
- Large – up to 12 GB
- Extra large – up to 36 GB
Note: When selecting an IDIR size, if your largest IDIR is not typical (not committed very often), then consider using the next smaller IDIR size.
Choose the target throughput (commits/hr) and use the corresponding minimum requirements for commitPoolThreads, processors, and memory available to the Coverity cloud system.
If you are using a non-default Commit Queue size higher than 80, adjust the system memory accordingly.