Hi @tomg10,
The easiest way to conserve startup memory is to reduce the number of LMS and CMS gunicorn workers via the EDXAPP_WORKERS ansible variable. This is assuming you installed Koa via the native installation procedure - is that correct? Once you set that variable, you’ll need to run the playbooks again.
The default is 4 * the number of cores for the LMS, and 2 * the number of cores for the CMS. For you, that means the default is lms: 8, and cms: 4. You may want to hard-code that to lms: 4 and cms: 2, and see if it helps.