Hello AV Colleagues,
One of our customer setup their new Enterprise cluster using Vertica 9.2 AMI. Initial tests using vcpuperf shows very high low load time in access of 400 (compared to high load times of 100+), suggesting CPU scaling being enabled.
I was wondering how do we handle this in AWS? Vertica documentation refers to our hardware guide, and the recommendations aren't very relevant to the scenario.They have already made recommended changes to max_cstate and intel_pstate in grub.conf. Not only that they can't access BIOS settings on AWS instances, but also that Vertica AMI should set everything as expected including CPU scaling settings.
Any guidance will be appreciated.
Regards,
Pravesh
