Bug 1933268

There are currently known issues related to `[workarounds]/disable_native_luksv1` and `[workarounds]/rbd_volume_local_attach` configuration options. These options are provided only as a temporary workaround for known performance regressions within libgcrypt and librdb. These workaround options will be removed once the regressions are resolved in the underlying RHEL release used by RHOSP. There are caveats associated with using either of these workaround options. Failure to adhere to these caveats can cause issues with RDB encrypted volumes. The caveats are as follows: - A support exception must be granted from CEE before enabling these workaround options in your environment. A support exception is required to allow Red Hat to track use of the workaround and to help you remove these workaround options in the future when they are disabled. - You must enable the workaround options across all Compute nodes in a given environment or host aggregate. - No move operations are supported between Compute nodes that have these workarounds enabled and those that do not have these workarounds enabled. - All existing instances on a Compute node that you want to enable these workarounds on must be stopped or migrated off the node before you enable the workarounds. You can restart the instances when the Compute service has been restarted with the workarounds enabled.