IBM WebSphere CloudBurst Appliance, version 2.0.0.3 -- Release notes

Release Notes for IBM WebSphere CloudBurst Appliance, Version 2.0.0.3

This document contains the release notes for the IBM WebSphere CloudBurst Appliance, Version 2.0.0.3.

Last updated 25Mar2011

Enhancements in WebSphere CloudBurst, version 2.0.0.3

In WebSphere CloudBurst, version 2.0.0.3, the following enhancements were made:

Support of VM:Secure for use with z/VM

Function:	VM:Secure enhances security and directory management for z/VM. Use VM:Secure to can control access to system resources, manage disk space and audit system activity. VM:Secure also enhances security by automatically maintaining synchronization between the z/VM user directory and the security rules.
Issue:	This function was added to enhance security and directory management for z/VM.

---

USB to serial connector cable driver

Function:

If you are installing a new appliance and you are using the provided USB to serial connector cable, the driver is available from Fix Central (http://www-933.ibm.com/support/fixcentral/). The driver is available in the usb_driver.zip file. The usb_driver.zip file contains the executable file to put the driver on your machine. This driver is only needed for new appliances using the USB to serial connector cable to connect to the console.

Issue:

If you have an older appliance already installed or if you are not using the USB to Serial connection for your console, you can ignore the usb_driver.zip file.

List of APARs fixed in WebSphere CloudBurst, version 2.0.0.3

APAR IC70300: In various charting functions in the web console, the final time stamp on the chart's horizontal axis can appear to be earlier than the requested ending time

Problem:

The final time stamp on the chart's horizontal axis can appear to be significantly earlier than the ending time requested by the user.  Although the chart IS showing all data within the starting/ending times requested by the user, the labels on the horizontal axis of the chart might stop before the requested ending time.  The charting algorithm on the appliance aggregates data to ensure that the display remains legible.  When data is aggregated, the EARLIEST time stamp from the aggregated data is displayed on the horizontal axis.  For the starting time, this corresponds to the starting time requested by the user; for the ending time, however, it can be substantially earlier, depending how much data the appliance has for the requested time period.  Note: When the appliance has no data for the beginning of the requested time period, it automatically scales the horizontal axis to display as much information as possible to the user.  That is, it appears to "zoom in" to the end of the requested time period rather than display empty space on the chart.

Resolution:

The charting logic has been updated so that the time stamp of the most recent data included in the chart is ALWAYS shown on the horizontal axis.

---

APAR IC72556: Capture of an extended z/VM WPS image fails

Problem:

When a z/VM WPS image is extended and recaptured, it fails. The status on the virtual image is set to "Failed to load the virtual image in the catalog." The WPS image adds a new disk to the image at address 10A. The rpm code was using decimal logic when looking for disks to capture. Therefore, it missed the 10A disk. Mirage did not recognize that the captured virtual machine data contained the 10A disk.

Resolution:

The capture detection logic was changed to check for hex numbers rather than decimal numbers in addresses.

---

APAR IC72713: When using the REST API or CLI to examine a pattern associated with a virtual system, the pattern attribute might be null or point to an incorrect pattern

Problem:

When the data for the virtual system is returned, the pattern attribute might be null or point to an incorrect pattern.  New appliances (with 2.0.0.3 firmware) do not exhibit this problem, but errors during pattern creation can cause internal counters on the appliance to get out of synch.  When this happens, all future patterns on the appliance exhibit this problem when they are used to construct virtual systems.  Note: This problem affects the REST API and CLI *only*.  The web console always display the correct pattern virtual system association.

Resolution:

The appliance firmware has been corrected so that the virtual system is always associated with the correct pattern.

---

APAR IC72901: WebSphere CloudBurst Troubleshooting -> Outbound Connection -> PING result icons not displaying

Problem:	If multiple PINGs were done in a quick succession (within 15 seconds) the success and failure icons were not displaying. This was caused because the success icon faded over a 15 second time frame. Therefore, another success or failure icon would not display if another PING was issued during this 15 second interval.
Resolution:	The code was fixed to show the PING result icon, as expected, and the icons no longer require 15 seconds to fade.

---

APAR IC73607: When very large amounts of virtual disk image data are processed on the appliances, the repository could get corrupted

Problem:

The appliance uses an internal repository to efficiently store the contents of virtual disk images.  The logic used to manage this repository contained an error that could cause it to corrupt the repository when a sufficiently large amount of virtual image data had been added with any of the following mechanisms: Adding large numbers of virtual images Adding virtual images that contain large amounts of data on their virtual disks Extending/capturing virtual images that contain large deltas from the original virtual image Note: It is the amount of data that has been ADDED that triggers the error, regardless of how much of that data was subsequently deleted from the appliance.  In all known cases, the appliance had processed well over 100GB of virtual image data when the error was observed.  The amount of virtual image data is cumulative since the appliance was first powered on -- firmware upgrades do not change the internal state of the repository. The corruption could occur as part of the repository cleanup that happens after a virtual image is deleted.  If the repository became corrupted the only way to recover the lost data was from an appliance backup taken before the corruption occurred.  Any operations that tried to use the corrupted data from the virtual disk repository (deployments or virtual image exports, for example), failed.  The corruption might not be immediately noticeable due to the caching algorithms used by the appliance.

Resolution:

The appliance firmware has been corrected so that the corruption will not occur, regardless of the amount of virtual disk data has been processed.

---

APAR IC73796: Unable to deploy new virtual systems. Tasks are queued, and virtual systems are queued during deployment.

Problem:

The appliance was unable to deploy new virtual systems because the local storage for the appliance was full.  The appliance could not clean up the internal cache, causing the disk to fill up.  This happens when an appliance has a number of different virtual images that were deployed on to un-cached hypervisors.  If the appliance deploys the same virtual image repeatedly, this problem does not occur.  This problem is only visible on the 2.0.0.2 firmware level.

Resolution:

Upgrading to the latest firmware of the appliance prevents this error.

APARs not fixed in WebSphere CloudBurst, version 2.0.0.3

APAR IC73223: IP address is shown as active when the virtual system that contained it is deleted (ID only fix)

Problem:	An IP address can remain in the active state when the virtual system that used the IP address is deleted.  This occurs when there is a failure during the virtual system delete, but the delete still completes.  The result of the failure is that the IP address state is not set to inactive.
Resolution:	No fix is currently available.