3.3. Red Hat Virtualization Batch Update 2 (ovirt-4.1.3)

With this update, the ability to unplug a lease from and plug another lease into a running virtual machine has been added using new APIs. This provides the ability to move a virtual machine lease from one storage domain to another so that the original storage domain can be placed into maintenance.

With this update, engine.log has been added to the logs collection. The records are now parsed and sent to the central metrics store. This allows the administrator to analyse the logs in a more simple and comfortable way. This means that the engine.log records are available for analysis in Kibana.

Previous versions of Red Hat Virtualization were able to negotiate encryption protocol up to TLSv1 for external provider communication. This update adds the the ability to negotiate encryption protocol up to TLSv1.2.

Note: The exact version used for communication depends on highest version available on the external provider target.

With this update, ovirt-engine-metrics can now configure collectd to connect to the Red Hat Virtualization Manager's PostgreSQL database and get information from it. Previously, making collectd connect to the PostgreSQL database did not work and was disabled by BZ#1436001.

With this update, GWT symbol maps, used to clarify obfuscated client-side (GWT UI) stack traces (/var/log/ovirt-engine/ui.log), are now part of the core Red Hat Virtualization Manager UI packages. The packages are installed by default when the Red Hat Virtualization Manager is installed or upgraded.

This update means that the user now has meaningful client-side stack traces in /var/log/ovirt-engine/ui.log without having to install additional packages, which saves time when analyzing UI-related issues. This update also reduces the symbol maps' size by zipping them. The Administration Portal UI uncompressed GWT symbol maps previously took ~1 GB of disk space. The zipped version consumes only 50 MB. The content of the zipped GWT symbol map files is streamed by the Manager as required and the content is not physically extracted to disk. After the Manager install or upgrade, the GWT symbol maps files use the following location:

/usr/share/ovirt-engine/gwt-symbols/webadmin/symbolMaps.zip

NOTE: The webadmin-portal-debuginfo package, which previously provided the GWT symbol map files, must now be manually removed prior to engine upgrade, or the upgrade will fail. (Because this package is optional, you only need to remove it if you installed it manually.)

The Python SDK now supports asynchronous requests and HTTP pipe-lining. Users can send requests asynchronously and wait for the response later in code. This makes it possible to send multiple requests using multiple connections or pipelined connections, and wait for the response later, which improves performance when fetching multiple objects from the API.

The self-hosted engine Manager virtual machine upgrade procedure from Red Hat Enterprise Virtualization 3.6 on Red Hat Enterprise Linux 6 to Red Hat Virtualization 4.0 on Red Hat Enterprise Linux 7 now always tries to connect to the Manager API as admin@internal, making it more flexible.

This update adds a new VDSM hook called vdsm-hook-vfio-mdev, which makes the host capable of working with mdev capable devices such as GPUs. The hook is automatically installed. The following prerequisites apply for a virtual machine to use an mdev instance:

1) The host must have a kernel that supports the mediated device, and it must have the device and the correct drivers.
2) The predefined property, 'mdev_type', must be set and correspond to one of the mdev types supported by the device.
3) The virtual machine must be pinned to the host(s) with the device(s).

The supported mdev_type values can be determined by checking the host on which the device is present (not visible in the engine) by querying vdsm (vdsm-client Host hostdevListByCaps) or running the following bash script:

for device in /sys/class/mdev_bus/*; do
  for mdev_type in $device/mdev_supported_types/*; do
    MDEV_TYPE=$(basename $mdev_type)
    DESCRIPTION=$(cat $mdev_type/description)
    echo "mdev_type: $MDEV_TYPE --- description: $DESCRIPTION";
  done;
done | sort | uniq

With this update, the user can now control the location of the temporary directory created by the engine-backup command, using the command line option --tmpdir=DIR and the variable $TMPDIR. Previously, the engine-backup command always used /tmp. If /tmp was full, or close to being full, it was not possible to tell the engine-backup command to use a different directory.

Red Hat Virtualization Host (RHVH) now supports NIST 800-53 partitioning requirements to improve security. RHVH uses a NIST 800-53 partition layout by default, and existing configurations will be changed on update.

This update makes it easier to debug a disk snapshot's live deletion failures because the initial state of the volume chain is logged.

When collecting SOS reports from hosts, chrony and systemd SOS plugins can collect information about time synchronization. In addition, a new option --time-only has been added to ovirt-log-collector allowing information about time differences to be gathered from the hosts without gathering full SOS reports, saving a considerable amount of time for the operation.

Previously, the self-hosted engine setup failed if the system clock was set to a time zone that was behind UTC during the installation. This is due to the fact that the Red Hat Virtualization Host generates VDSM certificates at the first boot and if the clock is incorrect, the chronyd or ntpd processes resynchronized the clock. This lead to an invalid certificate if the time zone was behind UTC.

Now, Red Hat Virtualization Host generates the certificates after the chronyd or ntpd processes and waits two seconds for the clock to synchronize.

Note that if the Red Hat Virtualization Host is configured after the installation, or if the NTP server is too slow, the self-hosted engine setup may fail.

This release introduces a new VM Portal as a Technology Preview. A "VM Portal" link is now available on the Red Hat Virtualization Welcome Page. The VM Portal provides the same functionality that is available in the Basic tab of the current User Portal (now deprecated).

Previously, the self-hosted engine setup failed if the system clock was set to a time zone that was behind UTC during the installation. This is due to the fact that the Red Hat Virtualization Host generates VDSM certificates at the first boot and if the clock is incorrect, the chronyd or ntpd processes resynchronized the clock. This lead to an invalid certificate if the time zone was behind UTC.

Now, Red Hat Virtualization Host generates the certificates after the chronyd or ntpd processes and waits two seconds for the clock to synchronize.

Note that if the Red Hat Virtualization Host is configured after the installation, or if the NTP server is too slow, the self-hosted engine setup may fail.

With this update, the migration of a virtual machine to a different cluster can no longer be invoked from the UI. Regular migration within a cluster remains unchanged.

The User Portal is now deprecated. From Red Hat Virtualization 4.2, it will no longer be available. A new VM Portal will be added, but certain features may not be present at the time of release. A technology preview of the VM Portal is available with the 4.1.3 release. Customers utilizing the Red Hat CloudForms Self Service user interface to manage virtual machines will not be affected.