SUSE OpenStack Cloud 8

Release Notes

These release notes are generic for all SUSE OpenStack Cloud 8 components. Some parts may not apply to a particular component.

Documentation can be found in the docu language directories on the media. Documentation (if installed) is available below the /usr/share/doc/ directory of an installed system. The latest documentation can also be found online at https://www.suse.com/documentation/cloud/.

Publication Date: 2018-04-24 , Version: @VERSION

1 SUSE OpenStack Cloud

Powered by OpenStack™, SUSE OpenStack Cloud is an open source enterprise cloud computing platform that enables easy deployment and seamless management of an Infrastructure-as-a-Service (IaaS) private cloud.

2 Support Statement for SUSE OpenStack Cloud

To receive support, customers need an appropriate subscription from SUSE for more information, see https://www.suse.com/support/ .

3 Major Changes in SUSE OpenStack Cloud 8

SUSE OpenStack Cloud 8 is a major update to SUSE OpenStack Cloud and comes with many new features, improvements and bug fixes. The following list highlights a selection of the major changes:

  • OpenStack has been updated to the Pike release, and the deployment framework has been updated accordingly to support new features. Check Pike (https://releases.openstack.org/pike/) and Ocata (https://releases.openstack.org/ocata/) to see which new features got added.

  • All nodes used for Administration and OpenStack are now using SUSE Linux Enterprise Server 12 SP3 as operating system.

  • SUSE OpenStack Cloud 8 integrates with an externally deployed SUSE Enterprise Storage cluster which requires a separate subscription .

  • The PostgreSQL database used by the OpenStack control plane has been replaced in favour of MariaDB. The MariaDB implementation comes with several advantages like Galera Clustering and SSL encryption for the Client to Server communication.

  • The RabbitMQ configuration has seen significant improvements, including use of native RabbitMQ clustering, increased performance through the use of HiPE (High Performance Erlang), and more settings for fine-tuning such as the heartbeat timeout for faster client reconnection.

  • SUSE OpenStack Cloud 8 is delivered with two options for Kubernetes deployment.

    • OpenStack Magnum that requires a user to bring their own image

    • Heat templates to deploy ready to run SUSE CaaS Platform images.

    SUSE CaaS Platform is the preferred alternative over OpenStack Magnum and comes with several advantages like Transactional Updates or an up-to-date Kubernetes version.

  • SUSE OpenStack Cloud 8 is delivered with the Cloud Lifecycle Manager .

  • The Cloud Lifecycle Manager framework comes with several highlights:

    • Template-based cloud definition with a built-in configuration checker.

    • Version controlling using a git repository which provides auditability.

    • Provides flexibility to modify cloud configurations including networking without having to redeploy the SUSE OpenStack Cloud.

    • Support for VPN as a Service for the OpenStack Networking Service (Neutron) has been added.

    • Support for the OpenStack networking load balancing Service (Octavia) has been added.

    • Support for the OpenStack disaster recovery Service (Freezer) has been added.

    • VMware NSX-V support has been added and supports NSX-V 6.3.

    • Cloud Lifecycle Manager comes with an UI based installer for cloud configuration and deployment. The installer UI is integrated with SUSE Manager as well as HPE OneView to support the capabilities of hardware discovery and interrogation. SUSE OpenStack Cloud 8 supports SUSE Manager 3.1 and OneView 3.0.

  • The Crowbar deployment framework also comes with several highlights:

    • Several OpenStack API endpoints are now using memcached as as caching mechanism to speed up requests.

    • The OpenStack Block Storage (Cinder), Dashboard (Horizon), Compute (Nova) and Authentication (Keystone) Services now offer the ability to fine tune the resource limit controls via the respective barclamp.

    • The OpenStack Block Storage (Cinder), Compute (Nova) and Networking (Neutron) Service now offer the ability to configure HA rate limiting via the respective barclamp.

    • The Block Storage service for OpenStack (Cinder) uses the multi backend option by default.

    • The OpenStack Orchestration Service (Heat) uses the convergence engine by default.

    • The OpenStack Monitoring Service (Monasca) now can monitor Ceph and Libvirt.

    • The OpenStack Monitoring Service (Monasca) can be configured to send all internal communication through specific networks only.

    • The OpenStack Networking Service (Neutron) does automatic LBaaS rescheduling to another host if the orginal host is not reachable.

    • The OpenStack Compute Service (Nova) now uses Cells v2 by default.

    • Several expert settings have been added, such as:

      • the ability to configure the Ceilometer API timeout;

      • the ability to configure Cinder NFS mount options;

      • the ability to configure the Cinder over-subscription ratio for the NetApp storage backend;

      • the ability to configure the Horizon secret key attribute;

      • the ability to disable Ironic auto cleanup which can be useful if the cleanup take a lot of time;

      • the ability to configure Swift Object Storage as Ironic Storage backend;

      • the ability to configure custom PXE parameters for Ironic;

      • the ability to configure the Hash algorithm for the Keystone password;

      • the ability to configure Monasca log agent attributes delay, elapsed time, meta data size and number of logs;

      • the ability to configure the Monasca InfluxDB retention policy;

      • the ability to configure Neutron VMware DVS ML2 support;

      • the ability to configure the Neutron L2 Population agent boot time;

      • the ability to configure the Nova image cache manager interval;

      • and many more!

    • Various improvements to high availability support have been included:

      • Stateless active/active services, such as OpenStack API and scheduler services, are not handled by Pacemaker anymore, but instead directly managed by systemd which will restart them in case of failure. This drastically reduces the complexity of the HA configuration, while keeping the benefits of the HA setup.

      • RabbitMQ is now using native clustering instead of an active/passive HA architecture. This removes the needs of shared storage for RabbitMQ, increases the performance and offers load balancing.

4 Technology Previews

Technology previews are packages, stacks, or features delivered by SUSE. These features are not supported. They may be functionally incomplete, unstable or in other ways not suitable for production use. They are mainly included for customer convenience and give customers a chance to test new technologies within an enterprise environment.

Whether a technology preview will be moved to a fully supported package later, depends on customer and market feedback. A technology preview does not automatically result in support at a later point in time. Technology previews could be dropped at any time and SUSE is not committed to provide a technology preview later in the product cycle.

Please, give your SUSE representative feedback, including your experience and use case.

SUSE OpenStack Cloud 8 ships with the following technology previews:

  • Data Processing Module for OpenStack (Sahara), and the respective Crowbar barclamp for deploying it.

  • EqualLogic driver for Cinder.

5 Deprecated and Removed Features

The following features are deprecated as of SUSE OpenStack Cloud 8:

  • The Cloud Lifecycle Manager is the new default deployment solution for new SUSE OpenStack Cloud deployments. Nevertheless Crowbar is still supported and available as separate media and channels called SUSE OpenStack Cloud Crowbar 8. The Crowbar based SUSE OpenStack Cloud flavor is supported until 2020-05-31.

  • The crowbar command line utility is deprecated in favor of the crowbarctl command line utility.

  • The deprecated keystonemiddleware.auth_token in-process token cache used by ceilometer services is replaced with a memcached server. When HA is used, several memcached servers are configured and used, one for each ceilometer-server node. The keystone token data stored in memcached is authenticated and encrypted using a generated secret key.

  • Ceilometer V2 API has been deprecated. Also publishing events has been turned off by setting source in events_pipeline.yaml to not match any event, since Gnocchi and Panko which are optional components are not available.

  • The Alarming service (Aodh) is no longer supported in this release in favor of the integrated Monitoring solution (Monasca).

  • The Application Catalog (Murano) is no longer supported in this release. It was included as technology preview in the previous release.

  • Database as a Service (Trove) is no longer supported in this release. It was included as technology preview in the previous release.

  • XEN as Compute Hypervisor has been deprecated and will not receive new functionality or improvements. XEN support will be removed in the next version of SUSE OpenStack Cloud.

The following features have been removed in SUSE OpenStack Cloud 8:

  • With Crowbar, new deployments cannot use the PostgreSQL database backend, as MariaDB is now the default database for SUSE OpenStack Cloud.

  • With Crowbar, new deployments can only deploy RabbitMQ with high availability by using native RabbitMQ clustering. This provides significant benefits over the old active/passive architecture.

  • With Crowbar, the Pacemaker barclamp cannot configure DRBD anymore in new deployments. The DRBD support was only used as an option to provide shared storage for PostgreSQL or the active/passive setup of RabbitMQ, and both of these use cases are deprecated now.

  • The Collector service (Ceilometer) is no longer supported in this release. The collector introduces lags in pushing data to backend. To optimize the architecture, Ceilometer Notification agent now directly publishes the data to the database via dispatcher.

  • The option to set up MongoDB as a Ceilometer backend was removed in this release. The MongoDB support was included as technology preview in previous releases.

  • The SUSE Linux Enterprise Server based Magnum Orchestation Image is no longer supported in this release. We recommend to use SUSE CaaS Platform as alternative; see Major Changes: SUSE CaaS Platform for details. If Magnum is preferred, we recommend to use images provided by OpenStack upstream, see https://docs.openstack.org/magnum/pike/user/ for details.

  • With Crowbar, the Ceph barclamp is no longer supported in this release. We recommend to deploy SUSE Enterprise Storage as external storage solution and connect it to SUSE OpenStack Cloud.

6 Upgrading to SUSE OpenStack Cloud 8

Upgrading to SUSE OpenStack Cloud 8 support will be delivered as a maintenance update in the near future. The upgrade will only be possible from and to the Crowbar based deployment framework. Upgrading to Cloud Lifecycle Manager will require a new installation of the product as the frameworks follow different philosophies.

Upgrading is supported from SUSE OpenStack Cloud 7, and requires all the latest updates to be applied. If running a previous version, please first upgrade to SUSE OpenStack Cloud 7.

The upgrade will be non-disruptive for the workloads if all prerequisites are met: high availability setup, enough compute resources, etc. This means that the instances running in OpenStack will keep running, will still have network connectivity and access to OpenStack resources such as volumes during the whole upgrade process. However, the OpenStack APIs and the OpenStack Dashboard will be turned off during the upgrade process, which may impact end users of the cloud.

If a non-disruptive upgrade is not possible due to unmet prerequisite, then the disruptive process can be used. In this mode, the whole OpenStack infrastructure will be turned off for the upgrade, and it is important to suspend all running instances during the upgrade. However, it is not necessary to do so at the beginning of the upgrade procedure, as this step can be postponed until after the Administration Server has been upgraded to SUSE OpenStack Cloud 8, in order to keep the downtime as short as possible.

The upgrade is done via a Web interface guiding you through the process. The process will generate a backup of the Administration Server as well as a dump of the OpenStack database. It is highly recommended to save these data to allow recovery, should the upgrade process go wrong.

The complete upgrade process is documented in the Deployment Guide, which can be found online at https://www.suse.com/documentation/cloud/.

7 Documentation and Other Information

  • Read the READMEs on the DVDs.

  • Get the detailed changelog information about a particular package from the RPM (with filename <FILENAME>):

    rpm --changelog -qp <FILENAME>.rpm
  • Check the ChangeLog file in the top level of DVD1 for a chronological log of all changes made to the updated packages.

  • Find more information in the docu directory of DVD1 of the SUSE OpenStack Cloud 8 DVDs. This directory includes PDF versions of the SUSE OpenStack Cloud documentation.

  • https://www.suse.com/documentation/cloud/ contains additional or updated documentation for SUSE OpenStack Cloud.

  • Visit https://www.suse.com/products/ for the latest product news from SUSE and https://www.suse.com/download-linux/source-code.html for additional information on the source code of SUSE Linux Enterprise products.

8 Limitations

  • The SUSE Linux Enterprise Server 12 SP3 nodes deployed through SUSE OpenStack Cloud are not compatible with the Public Cloud Module for SUSE Linux Enterprise Server 12 SP3, because SUSE OpenStack Cloud provides more recent versions of the OpenStack client tools.

  • The x86_64 architecture is the only supported architecture for the Administration Server and the nodes managed by SUSE OpenStack Cloud.

  • Chef, Ansible and Galera are only supported as part and in combination with SUSE OpenStack Cloud.

9 Known Issues

The following issues are known and will be fixed in a future maintenance update.

  • The upgrade from SUSE OpenStack Cloud 7 depends on a maintenance update for SUSE OpenStack Cloud 7 that is in the process of being released.

  • No pre-built image for Magnum , Sahara and Trove are shipped with SUSE OpenStack Cloud; cloud administrators are responsible for creating such images.

  • With Crowbar, the removal of barclamps from a node does not necessarily shut down associated services or remove associated packages. This means that you may well run into problems if moving barclamp roles from one node to another. Manual remediation may be required in these cases.

  • With Crowbar, using High Availability with multicast transport on Neutron L3 nodes is causing issues in some cases due to conflicts with the networking configuration required by Neutron. This can lead, in the worst case, to breakage of the High Availability cluster. It is advised to use the unicast transport (which is the default) for High Availability to avoid this.

  • Limited enablement for vCenter Hypervisor in SUSE OpenStack Cloud: Starting with Newton, OpenStack no longer supports vCenter integration without NSX setup, however the latter is not yet supported in SUSE OpenStack Cloud. This will be improved in a future maintenance update.

  • With Cloud Lifecycle Manager, its not possible to install SMT on the deployer if the deployer is sharing database with the OpenStack control plane. As a workaround, the systems need to be manually subscribed to an external SMT, as there is no automatic support for external SMT, or deployment needs to happen with deployer outside cloud based input models.

  • When using Cloud Lifecycle Manager using SMT on deployer, it is possible that Cloud Lifecycle Manager detects more than one flavor of Cloud Lifecycle Manager and subscribes to all available, potentially conflicting update channels. Before starting Cloud Lifecycle Manager make sure that SMT only mirrors the flavor that you're using.

  • With Cloud Lifecycle Manager, the workflows to update or upgrade an installation could still be subject to change. Prior executing any update or upgrade, we urge to refer to the most latest documentation prior executing any steps.

  • With Cloud Lifecycle Manager, the data plane development kit (DPDK) integration is not functional. As a workaround the driver needs to be loaded manually with

    modprobe --allow-unsupported igb_uio

    before deploying a DPDK node.

  • With Cloud Lifecycle Manager, the integration with an externally deployed SUSE Enterprise Storage is not functional.

  • With Cloud Lifecycle Manager, the OpenStack load balancing service (Octavia) does not automatically upload the guest image. As a workaround, the image (from /srv/tftpboot/files/openstack-octavia-amphora-image) can be uploaded manually to the OpenStack Image Service (Glance).

  • With Cloud Lifecycle Manager, WebSSO is not working due to the usage of the UUID token provider in the OpenStack Identity Service (Keystone).

  • With Cloud Lifecycle Manager, audit log files for Backup/Restore and Disaster Recovery Service (Freezer) are not populated.

  • With Cloud Lifecycle Manager, VMWare ESX together with ovsvapp is not functional.

  • With Cloud Lifecycle Manager, the Bare Metal Service (Ironic) has a known issue with the agent_ilo driver. If the user image takes up more space than what's left on the ramdisk's root partition, the user image needs to be remastered and the ramdisk size increased.

  • With Cloud Lifecycle Manager, adding a new node after the initial deployment might fail due to invalid repository metadata. As a workaround, the repository key for the PTF repository can be manually added via an Ansible Ad-Hoc command:

    ansible -m shell -a "sudo zypper -n --gpg-auto-import-keys ref -f PTF" list-of-new-nodes

    .

10 How to Obtain Source Code

This SUSE product includes materials licensed to SUSE under the GNU General Public License (GPL). The GPL requires SUSE to provide the source code that corresponds to the GPL-licensed material. The source code is available for download at https://www.suse.com/download-linux/source-code.html. Also, for up to three years after distribution of the SUSE product, upon request, SUSE will mail a copy of the source code. Requests should be sent by e-mail to mailto:sle_source_request@suse.com or as otherwise instructed at https://www.suse.com/download-linux/source-code.html. SUSE may charge a reasonable fee to recover distribution costs.

12 Colophon

Thanks for using SUSE OpenStack Cloud in your business.

The SUSE OpenStack Cloud Team.

Print this page