4.4. Red Hat Virtualization 4.3 Batch Update 3 (ovirt-4.3.6)

BZ#1660451

Previously, some virtual machine operations unnecessarily blocked other VM operations. This led to some problems with monitoring while shutting down large virtual machines. The current release fixes these issues: It relaxes some conditions for blocking virtual machine operations and makes blocking safer. This should reduce monitoring problems experienced in some scenarios.

BZ#1691760

Previously, trying to enable virtual functions (VFs) on a Broadcom network card generated a "Failed to change the number of virtual functions" error. The current release fixes this issue.

BZ#1710725

The same subnet was used in the local network and the libvirt network, and as a result, the network subnets collided. In this release, no subnet collisions occur.

BZ#1717954

Previously, when you used the oVirt config files to edit the libvirt.conf config file, running 'systemctl status collectd' failed to collect metrics data, instead displaying errors such as the following:

BZ#1733438

During installation or upgrade to Red Had Virtualization 4.3, engine-setup failed if the PKI Organization Name in the CA certificate included non-ASCII characters. In this release, the upgrade engine-setup process completes successfully.

BZ#1734346

Previously, backticks inside the bash 'test' command caused errors to be logged when the bash script was interpreted. The current release fixes this issue by using $() and quoting the output.

BZ#1734671

The Database field guest_cur_user_name in the vm_dynamic Database Table was limited to a size of 255 characters, which was not enough to accept a large number of users (for example, more than 100) to log in. In this release, there is no character limit for the guest_cur_user_name field.

BZ#1740498

Previously, while VDSM service was starting, the nwfilter configuration removed and recreated the 'vdsm-no-mac-spoofing' network filter definition to ensure the filter was up to date. If many vNICs were using the filter, that operation took longer than expected and caused the VDSM startup to timeout. The current release fixes this issue by updating the filter, which takes less than a second independent of how many vNICs are using the filter.

BZ#1744572

Previously, while powering off a virtual machine, the RHV Manager occasionally displayed an error message that the VDSM Get Host Statistics command failed due to an Internal JSON-RPC error: {'reason': '[Errno 19] vnet111 is not present in the system'}. This error occurred because an interface disappeared while the host statistics were being gathered. The current release fixes this issue by suppressing this error under these conditions.

BZ#1746718

Previously, the gluster fencing policy check failed due to a non-iterable object and threw an exception. The code also contained a minor typo. The current release fixes these issues.

BZ#1748395

Previously, v2v was reporting the wrong disk type when using the -o rhv and -o vdsm option; copying disks created by virt-v2v failed with: InvalidParameterException: Invalid parameter: 'DiskType=1'. Although the issue of assigning the wrong disk type was fixed in virt-v2v upstream, we must be able to copy disks created using the "broken" version of virt-v2v. The current release enables us to do this by adding support for that incorrect disk type, "1".

BZ#1750245

Previously, the ioprocess package shipped in both the engine (RHV Manager) and host channels. The current release removes the redundant ioprocess package from engine channel.

BZ#1751142

During host activation, the engine did not check the glusterd status. The gluster service was restarted even though the glusterd was not stopped during maintenance. In this release, the quorum is not lost upon host activation.

BZ#1753168

Previously, a snapshot disk that downloaded when it was attached to a backup Virtual Machine got locked due to a failure to teardown the disk. The current release fixes this error by skipping the disk teardown in case of a snapshot disk.

BZ#1652861

Some administrators want to avoid restarting a virtual machine after changing its settings. When a user reconfigures a virtual machine in the Administration portal, the portal displays a message that recommends restarting the virtual machine. However, in some cases, restarting is not necessary. In the current release, the documentation indicates whether restarting the virtual machine is optional or required for specific settings. For more information, see the appendices at the end of the Virtual Machine Management Guide.

BZ#1680498

A new text field has been added to the affinity group dialog to set priority. It can contain any real number, not just integers. Priority can also be set using a new parameter in the REST API. When a Virtual Machine is started or migrated, a host is selected, with the broken affinity groups having the lowest priority. If not all affinity groups can be satisfied, the groups with lower priority are broken first.

BZ#1712747

Red Had Virtualization virtual machines now support passing ignition configuration for any guest that supports it, such as RHCOS or FCOS. A guest RHCOS/FCOS guest OS can now be configured by passing ignition configuration using the UI or the API.

BZ#1753116

vdsm now requires a host kernel with fix for CVE-2019-14835

BZ#1753901

The current release provides an API to probe the block size of the underlying filesystem. The vdsm package needs this API to support 4k storage on gluster.

BZ#1734619

The current release rebuilds Red Hat Virtualization (4.3.6) for the current version of Red Hat Enterprise Linux 7.

BZ#1745961

Rebase package(s) to version: 1.8.2

BZ#1720879

A new Knowledgebase article is available in the Red Hat Customer Portal - "Deploying Metrics Store on an existing OpenShift Container Platform deployment". Use your Red Hat customer login credentials to view the article - https://access.redhat.com/articles/4309251