2.8. Technology Previews

Rally is a benchmarking tool that automates and unifies multi-node OpenStack deployment, cloud verification, benchmarking and profiling. It can be used as a basic tool for an OpenStack CI/CD system that would continuously improve its SLA, performance and stability. It consists of the following core components:

The Data Plane Development Kit (DPDK) consists of a set of libraries and user-space drivers for fast packet processing, enabling applications to perform their own packet processing directly to/from the NIC, delivering up to wire speed performance for certain use cases. In addition, OVS+DPDK significantly improves the performance of Open vSwitch while maintaining its core functionality. It enables the packet switching from the host’s physical NIC to the application in the guest instance (and between guest instances) to be handled almost entirely in user-space.

In this release, the OpenStack Networking (neutron) OVS plugin was updated to support OVS+DPDK back end configuration. OpenStack projects can now use the neutron API to provision networks, subnets and other networking constructs, while using OVS+DPDK to gain improved network performance for instances.

Red Hat OpenStack Platform 8 now includes a technology preview of integration with the OpenDaylight SDN controller. OpenDaylight is a flexible, modular, and open SDN platform that supports many different applications. The OpenDaylight distribution included with Red Hat OpenStack Platform 8 is limited to the modules required to support OpenStack deployments using OVSDB NetVirt, and is based on the upstream Beryllium version. The following packages provide the Technology Preview: opendaylight, networking-odl

Integration of real time KVM with the Compute service further enhances the vCPU scheduling guarantees that CPU pinning provides by reducing the impact of CPU latency resulting from causes such as kernel tasks running on host CPUs. This functionality is crucial to workloads such as network functions virtualization (NFV), where reducing CPU latency is highly important.

The Red Hat OpenStack Platform director has the ability to integrate services from OpenStack's containerization project (kolla) into the Overcloud's Compute nodes. This includes creating Compute nodes that use Red Hat Enterprise Linux Atomic Host as a base operating system and individual containers to run different OpenStack services.

OpenStack Compute includes the concept of Cells, provided by the nova-cells package, for dividing computing resources. For more information about Cells, see Schedule Hosts and Cells.

Alternatively, Red Hat Enterprise Linux OpenStack Platform also provides fully supported methods for dividing compute resources in Red Hat Enterprise Linux OpenStack Platform; namely, Regions, Availability Zones, and Host Aggregates. For more information, see Manage Host Aggregates.

OpenStack Database-as-a-Service allows users to easily provision single-tenant databases within OpenStack Compute instances. The Database-as-a-Service framework allows users to bypass much of the traditional administrative overhead involved in deploying, using, managing, monitoring, and scaling databases.

Distributed Virtual Routing (DVR) allows you to place L3 Routers directly on Compute nodes. As a result, instance traffic is directed between the Compute nodes (East-West) without first requiring routing through a Network node. Instances without floating IP addresses still route SNAT traffic through the Network node.

Red Hat OpenStack Platform 8 includes a Technology Preview of DNS-as-a-Service (DNSaaS), also known as Designate. DNSaaS includes a REST API for domain and record management, is multi-tenanted, and integrates with OpenStack Identity Service (keystone) for authentication. DNSaaS includes a framework for integration with Compute (nova) and OpenStack Networking (neutron) notifications, allowing auto-generated DNS records. In addition, DNSaaS includes integration support for PowerDNS and Bind9.

The Object Storage service includes an EC storage policy type for devices with massive amounts of data that are infrequently accessed. The EC storage policy uses its own ring and configurable set of parameters designed to maintain data availability while reducing cost and storage requirements (by requiring about half of the capacity of triple-replication). Because EC requires more CPU and network resources, implementing EC as a policy allows you to isolate all the storage devices associated with your cluster's EC capability.

The OpenStack File Share Service provides a seamless and easy way to provision and manage shared file systems in OpenStack. These shared file systems can then be used (mounted) securely to instances. The File Share Service also allows for robust administration of provisioned shares, providing the means to set quotas, configure access, create snapshots, and perform other useful admin tasks.

The following section briefly describes the new features included in the File Share Service for Red Hat OpenStack Platform 8.

Manila Horizon dashboard plug-in

With this release, users can now interact with the capabilities that the File Share Service provides via the dashboard, including an interactive menu for creating and working with shares.

Share migration

Share migration is a new feature that enables migration of shares from back end to back end.

The following approaches are available:

For the second to work, every driver during server_setup method must create a port that allows connectivity between the share server and manila node.

Availability zones

The File Share Service client's share creation code now accepts and uses availability zone arguments. This also allows the preservation of availability zone information when creating a share from a snapshot.

Oversubscription in thin provisioning

This release adds support for oversubscription in thin provisioning, which addresses the use case where certain drivers still report infinite or unknown for their capacity, potentially leading to oversubscription. This update adds the following parameters:

The Firewall-as-a-Service plug-in adds perimeter firewall management to OpenStack Networking (neutron). FWaaS uses iptables to apply firewall policy to all virtual routers within a project, and supports one firewall policy and logical firewall instance per project. FWaaS operates at the perimeter by filtering traffic at the OpenStack Networking (neutron) router. This distinguishes it from security groups, which operate at the instance level.

Operational Tools are logging and monitoring tools which facilitate troubleshooting. With a centralized, easy-to-use analytics and search dashboard, troubleshooting is simplified, and features such as service availability checking, threshold alarm management, and collecting and presenting data using graphs are available.

VPN-as-a-Service allows you to create and manage VPN connections in OpenStack.

Time-Series-Database-as-a-Service (gnocchi) is a multi-tenant, metrics and resource database. It is designed to store metrics at a very large scale while providing access to metrics and resources information to operators and users.