Release Notes #

The documentation for this product has been extended. Some manuals are only available online.

The new Installation and Setup Quick Start guides you through the setup of a very basic two-node cluster.

The Geo Clustering Quick Start explains how to set up an example scenario with a two-site Geo cluster, one arbitrator, and data replication via DRBD.

Both Quick Starts are available on the installation media and online.

The comprehensive Administration Guide has been removed from the installation media, but is available online. Find the complete documentation for this product at https://www.suse.com/documentation/sle-ha-12/ (https://www.suse.com/documentation/sle-ha-12/).

As with resource agents, the cluster calls an external program, an "alert agent", to handle alerts. The cluster passes information about the event to the agent via environment variables. Agents can do anything desired with this information, for example, they can send an e-mail, log to a file, or update a monitoring system.

Multicast is not available for clusters running on Amazon Web Services.

The ha-cluster-bootstrap scripts now detect if a node is an AWS machine. In that case, the cluster is automatically configured for unicast communication.

It is now possible to bootstrap a cluster with unicast instead of multicast. To configure unicast as transport protocol for cluster communication, use:

ha-cluster-init -u

In SLEHA 12 SP1, Hawk 2 already was the default for new installations. However, when upgrading from a previous version of SLEHA, Hawk 1 would be used.

In SLEHA 12 SP2, Hawk 2 replaces Hawk 1 even for system upgrades.

For configuring ACL rules, it can be helpful to be able to view the text configuration directly in Hawk, as well as the XML format configuration.

The Edit Configuration page now has subpages displaying the crmsh configuration and the raw XML configuration.

With the previous Simulator, it was not possible to see a summary of changes included in the simulation or to apply the changes to the live configuration.

The new Batch Mode simulation interface of Hawk2 shows the configuration differences compared with the live configuration as well as any simulated events.

When enabling Batch Mode, a control panel appears at the top of the Web interface, allowing you to inject simulator events. In this mode, the simulator runs on any configuration changes or event injections.

You can view the detailed summary of the current simulation and see the list of changes that have been made since entering batch mode.

You can either apply the simulated changes to the cluster configuration or cancel them.

In the past, Hawk displayed utilization attributes (defining a node's capacity) and the current capacity consumed by resources running on the node. However, editing utilization attributes had to be done via the command line.

Now Hawk allows you to edit utilization attributes for both for nodes and resources: You can define the capacity a certain node provides and the capacity a certain resource requires.

Configuring fencing is crucial for a fully functional cluster.

Hawk2 now has a wizard that allows configuring fencing with SBD.

Some operations were not available from the Hawk2 interface, such as renaming resources.

You can now rename resources directly from the Hawk2 interface. In addition, the following operations are now also available for groups:

Previously, ldirectord did not support HTTP checks for IPv6.

This issue was fixed by adding the packages perl-Net-INET6Glue and perl-IO-Socket-IP as dependencies to ldirectord.

HAProxy 1.6 has a lot of new features and significantly improves performance over 1.5.

New features in HAProxy 1.6 include:

The performance has also been improved a lot with:

The great flexibility offered by this version will allow many users to significantly simplify their configurations. Some users will notice a huge performance boost after they enable the features designed for them.

Hawk 1 lacks support for new features in Pacemaker and its look and feel does not match that of other SUSE products.

Hawk 1 has been dropped and is completely replaced by Hawk 2.

With Pacemaker remote nodes, it is possible to have a node and a resource with the same name. In previous versions SUSE Linux Enterprise High Availability Extension, crm configure show could not filter for nodes only or resources only.

crm configure show now supports the operators and and or which allow combining name and type filters. To show only the node webui, use:

crm configure show type:node and webui

To show all location or colocation constraints, use:

crm configure show type:location or type:colocation

Sometimes, it is desirable to replace a configuration completely.

Using the command crm configure load push to add and remove configuration elements in a single command. For example, use:

crm configure load push <new-configuration>

In past versions, Relax-and-Recover (Rear) did not support disaster recovery on UEFI systems.

SLEHA 12 SP2 ships with Relax-and-Recover 1.18 and the new helper tool ebiso to allow for disaster recovery on UEFI systems.

ebiso can be used to make a UEFI-bootable Relax-and-Recover recovery system ISO image.

To set this up, install the package ebiso and add the following line to /etc/rear/local.conf:

ISO_MKISOFS_BIN=/usr/bin/ebiso

Auto-Promotion is enabled as a tech preview to help customers to work with the new capabilities of DRBD9. This feature might be fully supported in future version.

So far, the resource agent ocf:linbit:drbd could only be used with DRBD 8.

DRBD 9 now supports the resource agent ocf:linbit:drbd. You can either continue to use this resource agent to control DRBD or instead switch to file system resources that can promote the underlying DRBD device on demand, using the auto promote feature of DRBD 9.

Between DRBD 8 (shipped with SLE HA 12 SP1) and DRBD 9 (shipped with SLE HA 12 SP2), the metadata format has changed. DRBD 9 does not automatically convert previous metadata files to the new format.

After migrating to 12 SP2 and before starting DRBD, convert the DRBD metadata to the version 9 format manually. To do so, use:

drbdadm create-md "RESOURCE"

SLEHA 12 SP2 now ships with DRBD 9, the successor to DRBD 8.

With DRBD 9, SUSE supports the same use cases that were also supported with DRBD 8. Use cases beyond that, such as setups with more than 2 nodes, are not supported.