Please read the following release notes carefully, before continuing. Note that when installing this update on a Collax Inter-V cluster the following steps have to be performed:
Put the first node into maintenance mode. Go to Status/Maintenance → Cluster Nodes and execute the Standby action. If successful please stop the service "HA Cluster" in the form Status/Maintenance → Cluster Services.
Go to the local administration of the node and access the form System Operation → Software → Systemupdate. Then click Get Package List. This will download the list of available update packages. The successful update of the package list is indicated by the message Done!.
Click Get Packages. If the packages are downloaded successfully then the message Done! will be displayed.
Click Install. This action installs all update packages. The end of this process is indicated by the message Done!.
A new kernel is going to be installed. The system will be rebooted automatically after installing the update. An appropriate note will be shown if the update process is completed.
After the reboot resume the first Node as Cluster member: Go to the Cluster administration form Status/Maintenance → Cluster services and start the service "HA cluster" of this nodes. Change after that Status/maintenance → Cluster Nodes and resume the node as Cluster member by using the action Set active.
Next check the status of the DRBD disk synchronization in the form Shared Storage → Embedded SAN. Do not continue with the next step until 100% is reached in the column "Sync" for all DRBD disks.
Please repeat this procedure on the next node that needs to be updated.
To install this update on a Collax V-Cube without a Cluster interconnection please follow these steps:
Please read this document before proceeding to the next step.
In the administration interface go to System → System Operation → Software → System Update and then click Get Package List. This will then download the listed update packages. The successful update of the package list is indicated by the message "Done!".
Click Get Packages to download the software packages.
Click Install. This action installs the update. The end of this process is indicated by the message Done!.
A new kernel is going to be installed. The system will be rebooted automatically after installing the update. An appropriate note will be shown once the update process is completed.
New in this Release
Problems Fixed in this Release
Known Issues
A new kernel is going to be installed and a reboot of the system is necessary. The output of the progress bar may stop before rebooting.
Note: Please wait, until all software packages are installed. The reboot of the system will then be initialized automatically and the server is going to be available after a few minutes.
An update from V-Cube 6.0.6 conditions the installation of V-Cube 6.0.8. It is necessary to perform this intermediate step to assure the operation of the virtual machines during the update. The intermediate step is automatically enforced by the Collax update server.
If cluster nodes shall be updated it is necessary to update all nodes to V-Cube 6.0.8 first. After all nodes are updated to Version 6.0.8 the Version 6.0.10 shall be installed.
When creating snapshots, the virtual machine used to be briefly suspended for the duration of the backup of the memory content. As of this update, the memory content can backed up live, without interrupting the operation. The live snapshot option greatly increases the availability of the virtual machine to users during operationpresent snapshots. This is resolved with this update.
The new option is available in the Snapshots dialog.
Collax V-Cube opens disk image files and snapshots as normal files, which results in them being cached by the V-Cube like any other file. If disk image files and snapshots (also backup) are used the cache mode is to be set correctly from this update on. On principle the setting of the HDD caching mode depends on the filesystem used within the virtual guest. For File systems like NTFS, ext4 and XFS that support barrier by default the mode "Performance" is to be set. The file systems FAT/VFAT and ext3, which do not support barrier by default, the value "Data security" needs to be set. The Setting HDD-Caching is found within the dialog for configuring a virtual machine.
From this version on the action Remove Cluster filesystem is available for drives (Embedded SAN- or iSCSI-drive) not being used as cluster share anymore. This is useful to provide these drives, which prior were used as cluster share, as disk again within virtual machines.
From this version the console of virtual maschines is also going to be exported via remote desktop protocol. The activation and configuration of permissions is embedded into the configuration of the VNC console. Access via RDP can also be protected by a password. The user only needs the conventional Windows(tm) RDP client pointing to a cluster node which acts as RDP server. Afterwards a list of all accessible virtual machines is shown.
With this update, the V-recovery function set is made available for quick access to backed-up virtual machines. The function scope enables quick access to the file system of backed-up machines. This File Recovery can make the file system of the selected machine available in the network via FTP. In this way, individual files can be restored from the backup data of the virtual machine without restoring the original VM. Supported file systems: VFAT, NTFS, Linux Ext, ReiserFS, XFS. File systems in dynamic disks (SFS) or Logical Volumes (LV) are not supported.
Alternatively, the entire virtual machine can be booted temporarily from the backed-up data without affecting the original VM or the operating environment. This subfunction is called Instant VM.
The third function that V-recovery supports is the verification of the backup data for completeness and integrity. This function can be used to ensure that the backed-up data are correct and can be restored and that the virtual machine including snapshots can be booted if necessary.
To facilitate the initial setup, the registration wizard now enables the automatic installation of all licensed modules immediately after the successful registration. This speeds up the initial setup considerably.
Version 6.0.8 introduced an automatic logout from the system. As of this update, you can modify the idle time of the administration session under Settings → Usage Policy → Environment → Administrator (the default setting is one hour).
Besides the information about the disk delta, machine state or online/offline mode additional hardware information about CPU and RAM is going to be shown from this update on in the list of snapshots of a virtual machine.
A watchdog timer is a component in a system that monitors other system components that contact the watchdog timer at certain intervals (dead man's switch). In the case of an error, the watchdog triggers a predefined action to eliminate the error. As of this update, the Intel® 6300ESB Watchdog Timer is available to Collax V-Cube virtual machines. This watchdog timer can be initialized by virtual machines in order to confirm the correct system operation at certain intervals. If the signal is no longer transmitted to the watchdog timer, the defined action (e.g. reset) will be executed.
By default, all services of the involved nodes are actively monitored in a Collax cluster array. From this update, VM-side services can also be monitored. On the one hand, the test spectrum comprises several network-based services like DNS, HTTP, POP3, and SMTP. On the other hand, system-internal functions and values such as CPU, RAM, swap, processes, running services, events, and hard disks of the virtual machines can also be monitored. A detailed description of the use and configuration is provided at www.collax.com.
The ease of use of the Collax administration GUI is based among other things on triggering background processes which take away a lot of the work from the operator. From this update, an interface is implemented into the administration GUI which visualizes background processes including their result.
Among other things, the cluster management is based on the administration of resources, which are described with an actual state and a target state. This update activates a test that monitors these states. If the actual state corresponds to the target state, the result is OK. If the states do not match, the situation will be classified as CRITICAL, meaning that there is an error. In the case of an error, the system administrator is automatically informed by e-mail.
The following important system packages are installed with the new version:
Replicating block device driver and tools drbd 8.3.11-ak33
Corosync Cluster Engine 1.4.1-ak17
kernel 2.6.32.45-ak83
If the hard disk of a virtual machine of which snapshots had been generated was changed, e.g. if it was resized, and if the snapshots were subsequently merged with the base image, the file system of the virtual machine could be defective. To avoid this error, this update prevents the resizing of the hard disk in such a case.
Previously, joining a cluster could fail if there were different jumbo frame settings or different network objects with the same name. With this update, these items are verified before joining the cluster, and the user is provided with information on how the problem can be solved. In this way, it is made sure that the node can duly join the cluster.
Previously, the cluster domain could be selected independently from the FQDN. To prevent the cluster domain from being different from the FQDN, which would cause authentication problems between the nodes, the cluster domain can no longer be selected freely from this update on. Rather, it is set automatically according to the FQDN set for the node. The pseudo TLD .local won't be allowed any more. Additionally, the integrity of the DNS, FQDN, cluster domain and software version of a cluster node is verified when joining a cluster.
The backup of virtual machines also requires the backup of the virtual hard disks, for which a lot of storage space is needed on the backup medium. From this release on, the incremental backup method can be selected in order to reduce the need for storage space when backing up virtual machines. For the incremental backup, only the data of the virtual machine that changed since the last backup are stored in the form of a snapshot. The option can be found under Settings → Virtualization → General. Incremental backups are now performed by default if the backup type is duly configured.
To back up a virtual machine without interrupting the operation or to perform incremental backups, a snapshot is generated prior to the backup. The snapshot is stored in the vmData share or in the cluster share, depending on whether it is a local VM or a VM in the cluster. Before the backup, it was previously not checked whether the storage space was sufficient for the maximum size of the snapshot. From this update on, the free storage space is checked prior to the backup, and the process is canceled in case the storage space is insufficient.
Up to now ISOs, snapshots, Collax templates and diskimage files located on the cluster share were written to backup media, when the cluster share was selected for data backup. Since snapshots and diskimage files are also included in the data backup of a virtual machines, data was written redundantly and more space was used than necessary. From this update on ISOs and Collax templates can be chosen separately for data backup, whereby less storage space is required.
Previously, the table row focus did not work correctly in tables displayed in a section in the dialog with scroll bars. This error only occurred in connection with scroll bars and the Firefox browser. From this update on, the qx.html.Scroll method is deactivated in the AJAX framework; thus, the table row focus works correctly in new browsers.
When scrolling in dialogs and accessing pop-up objects such as list boxes or context menus, these pop-up objects were not displayed in the correct place. This update corrects the qx.ui.popup.Popup method, ensuring correct display of all pop-up objects in dialogs.
Under certain circumstances, the system information graphs of virtual machines were not displayed from Collax V-Cube version 6.0.8 on. This was caused by the process name of VMs, which had the prefix "qemu:" from version 6.0.8 onward. For this reason, the directories of the statistical records could not be accessed via the GUI. This error has been eliminated as of this update; the graphical system information for virtual machines is displayed correctly, and the previous records are retained.
Previously, it was not possible to specify more than 16 virtual CPUs via the administration interface, although 64 CPUs are possible per VM. This limitation has been eliminated as of this update.
If virtual machines use contents such as ISO images, snapshots, or disk image files of the cluster share and the cluster share is resized, these VMs will be stopped automatically and subsequently rebooted.
If a paused VM is migrated, it will be in the "running" state after the migration, even though the VM control continues to display it as "paused" . The display of the operating state can be corrected by executing the action "Resume" .
If Windows Server 2008 R2 is to be installed on Virtio hard disks using the Virtio driver CD for Windows, Windows reports the error code 0x80300001. The error message occurs if the partitions are recognized and if the user clicks the button Next. The message means, that the installation CD of Windows needs to be inserted again.