Release Notes Main Page | Introduction | What's New and Improved | Known Issues | Resolved Issues

Resolved Issues

Issues resolved from ESX Server RC1 and VirtualCenter Beta Refresh with this release of RC 2.

  • Virtual Infrastructure Client 2.0 Does Not Upgrade VirtualCenter Client 1.x

    Description: Though the installation of the Virtual Infrastructure Client 2.0 appears to complete successfully, when attempting to use the VI Client to login to an ESX Server host results in the message "Error Connecting - One or more of the types in the assembly unable to load."

  • SwapFile.vswp Not Removed During Upgrade from ESX Server 2.x to ESX Server 3.0

    Description: ESX Server 2.x has a global swap file "SwapFile.vswp" in VMFS2. After upgrading ESX Server 2.x to ESX Server 3.0 the file is neither needed by ESX Server 3.0, nor removed.

  • SwapFile.vswp Not Removed During Upgrade from ESX Server 2.x to ESX Server 3.0
    Description: ESX Server 2.x has a global swap file "SwapFile.vswp" in VMFS2. After upgrading ESX Server 2.x to ESX Server 3.0 the file is neither needed by ESX Server 3.0, nor removed.

  • Unable to Change Disk Mode of Legacy Virtual Machines (VM2) running on ESX Server 3.0 Datastore (VMFS3)

    Description: If you try to change the disk mode of a Legacy "persistent" mode virtual machines running on a ESX Server 3.0 Host to "non-persistent" or "append" or "undoable" diskmodes using the Virtual Machine Configuration editor. VirtualCenter displays the error: vim.fault.InvalidDeviceBacking-A device is being Added/Edited with an incompatible or invalid device backing

  • Virtual Machine with Raw Disks (RDMs) Attached Fails to Power On After Host Is Upgraded from ESX Server 2.x to ESX Server 3.0

    Description: When a host is upgraded from ESX Server 2.5 to ESX Server 3.0, if the host has virtual machines that had raw disks (RDMs) attached, any such virtual machine will fail to power on after the host upgrade is complete.

  • Upgrade Fails with Purple Screen If SAN Disconnects During VMFS2 to VMFS3 Upgrade

    Description: If you are performing the automatic upgrade from VMFS2 to VMFS3, and the ESX Server host loses connection with the LUN because of a cable pull, SCSI connectivity issue, or similar problem, the system fails with a purple screen.

  • Purple Screen If Disconnecting From SAN During VMFS2=>VMFS3 Upgrade

    Description: If you are performing the automatic upgrade from VMFS2 to VMFS3, and you lose connection with the LUN because of a cable pull, SCSI connectivity issue, or similar problem, the system fails with a purple screen.

  • Problems with Legacy Windows 2003 Guest Operating Systems

    Description: If a Windows 2003 virtual machine that was created on an older ESX Server is moved to ESX Server 3.0, either through a migration or a host upgrade, certain problems can arise. Specifically, although the virtual machine can be powered on, editing the configuration or cloning the virtual machine might fail. ONly Windows 2003 guest operating systems are affected.

  • Consolidated Backup doesn't Work for RDMs in Logical Compatibility Mode

    Description: Do not use Virtual Machine Consolidated Backup (VCB) to back up virtual machines with virtual disks using Raw Device Mappings (RDMs) in logical compatibility mode (non-passthrough RDMs).

    Trying to create snapshots of the virtual machines using RDMs might result in data corruption.

  • Snapshots and VMware Consolidated Backup Not Supported in RC for Virtual Machines with RDMs

    Description: Using virtual machines with one or more virtual disks using RDMs can lead to data corruption if the virtual machine is a snapshot or uses VMware Consolidated Backup. This issue applies to both pass-through and non-passthrough RDMs. This issue does not apply if the RDMs are use in the default persistent mode with no snapshots.

  • PORT Variable in VMware Consolidated Backup config.js is Ignored

    Description: Using the PORT variable in the config.js file of the VMware Consolidated Backup does not work.

  • Template Upgrade on a Shared LUN Does Not Complete

    Description: If you attempt to upgrade a template on a LUN which is shared between an ESX Server 2.5.x host and an ESX Server 3.0 host, the upgrade does not complete.

  • Guest Operating System Detects New USB Hardware after Migration
    Description: The guest operating system can detect an unchanged USB configuration as new hardware when older virtual machines are migrated to an ESX Server version 3 host. This usually occurs following one of these activities:
    • You power on the virtual machine in Legacy mode.
    • You upgrade the virtual hardware and then power on the virtual machine.
      The new USB controller is permanently added to the virtual machine.

  • Installer Incorrectly Selects Red Hat LUN for Upgrade

    Description: The ESX Server installer might mistakenly select LUNs containing Red Hat Linux installations for upgrade, including disks used for non-passthrough Raw Device Map (RDM) virtual machines. This can result in upgrade failure and potentially cause the Red Hat Linux installation to stop working. One cause of this error is that a Linux virtual machine using RDMs exists on the ESX Server host being upgraded. This error is rare, but VMware strongly recommends employing the workaround described below in order to avoid damage to existing Red Hat Linux installations.

  • VMware Tools Auto-Upgrader Does Not Work for Windows Guest Operating Systems

    Description: When you run the auto-upgrader, the script remains in a hanged state after displaying /New Datacenter/<virtual machinename>: tools update completed successfully. The virtual machine remains powered on.

  • VirtualCenter Does Not Override a Host's License Source When Host Is Added

    Description: When a host is added to the VirtualCenter inventory, and the host is configured to use a different license source than that of VirtualCenter, VirtualCenter does not override the host license source with its own, even when it is configured to do so. VirtualCenter's default configuration is to override the host license source.

  • H/W iSCSI: DHCP is Not Supported Description: When using HW initiator iSCSI, the IP network must be reliable. When there is packet loss on the network, the network slows down and the ESX Server gets too many SCSI aborts and the Virtual machines no longer function. This packet loss occurs when multiple physical hosts use the same iSCSI target and there is enough traffic to fill the buffer on the switch.

  • Hardware iSCSI: Masking/Preferred Path Inconsistencies When the Path Policy Is Fixed

    Description: This is visible only you use esxcfg-mpath in the command line. If the path policy is fixed for a particular lun, when we mask/unmask the LUN for a particular vmhba, it does not fall back to the original HBA.

  • Virtual Machine Reports Error: "@BlueScreen: Exception type 14 in world 1024:console @ 0x8fcca2"

    Description: Changing the VMKernel's TCP/IP stack IP address during NFS I/0, the I/O hangs for 45 seconds. The virtual machine might report the error:

    @BlueScreen: Exception type 14 in world 1024:console @ 0x8fcca2
    0xc02ccc:[0x8fcca2]lpfc_sli_ringtxcmpl_get+0x6e(0x3c4bf030, 0x3c4bf128, 0x3c571030)
    0xc02cfc:[0x8faf5e]lpfc_sli_process_sol_iocb+0x2e(0x3c4bf030, 0x3c4bf128, 0x3c571030)

  • <Service Console Commands ls and rm Result in a Failure

    Description: Due to a stale directory entry for a file, the user sees error messages, such as
    ls
    /vmfs/volumes/cx500-10g/autotest-runtime: No such file or directory
    or
    rm
    cannot remove directory
    `autotest-runtime/win2000ServWithSP4.Aug_10_2005.SAVED.6': Directory not empty

    This happens when there is an error with the removal of the directory entry. An entry can persist, and the user can no longer remove it.

  • Nessus Issues a Warning Message When CIM Is Running

    Description: Nessus issues the following message when CIM is running:

    Warning found on port unknown (5989/tcp)
    Your web server seems to accept unlimited requests. It might be vulnerable to the 'WWW infinite request'
    attack, which allows a cracker to consume all available memory on your system.

    *** Note that Nessus was unable to crash the web server
    *** so this might be a false alert.

  • Firewall Ports Need to be Opened Before Installing Backup Agents on ESX Server 3.0

    Description: ESX Server 3.0 does not automatically open the firewall ports required for third-party backup agents. As a result, backup agents are unable to communicate with the ESX Server host.

  • Unable to Create Virtual Machines on NFS Volumes Configured as NAS Datastores

    Description: The NFS volume was published with restrictions on root access (root squash enabled). In order to be fully functional as a NAS datastore, VMware system software, such as VirtualCenter, esx, and the connection agent running on the esx, must be able to manipulate the file system to create directories and set permissions. Many NFS servers do not allow full root privileges to remote users. As a result, when a remote user like VirtualCenter or an ESX Server host tries to create virtual machines on an NFS datastore, the attempt fails with a version of the following message:

       vmodl.fault.InternalError -- An internal error occurred on the server

  • VMFS3 Volume Labels Created Using Symbolic Links Do Not Display in the VI Client Datastore Browser

    Description: After adding or changing a VMFS3 volume label using the ln -sf command, the VI Client does not display the volume label in the Datastore Browser; it displays only the UUID of the volume.

  • Unable to Create Virtual Machines Through VirtualCenter on NFS Volumes Published with Root Access Restrictions

    Description: The NFS volume was exported with restrictions on root access (root squash enabled). Because VirtualCenter access NFS datastores using the user ID (uid) and group ID (gid) for root, you need to either:

    • Turn off root squash.
    • Enable the user root squash redirects to (typically nfsnobody) to create and delete directories on the mounted NFS datastore.

    If the NFS datastore is not configured to meet one of these two conditions, your attempt to create a virtual machine fails with a version of the following message:

    vmodl.fault.InternalError -- An internal error occurred on the server

    This problem occurs only if you are managing the ESX Server host through VirtualCenter, which does not allow you to assign a delegate user (a user other than root) to access storage. If you manage the host exclusively through a direct connection with the VI Client, you can assign a delegate user and access NFS storage through that user.

  • When You Change Hardware on an ESX Server Host, the Host Fails to Boot or Boots With No VMKernel Loaded.

    Description: The automatic hardware change detection mechanisms incorporated in ESX Server do not recognize certain hardware configuration changes. Without manual intervention, the hardware change is not recognized unless you reinstall ESX Server with the hardware in place.

  • Adapter Model Information Not Displayed Correctly in Client

    Description: VI Client and Web Access Client do not display the complete model information for a storage adapter correctly.

  • Must Create VMotion/NAS vswitch Before Using NAS

    Description: You must use the client to create a VMotions/NAS vSwitch before you can use NAS.

  • VMFS Upgrade and Re-Layout Currently Two Steps

    Description: You currently have to first upgrade VMFS, then re-layour VMFS. In the future, clicking the Upgrade VMFS button will automatically perform re-layout.

  • Software iSCSI Not Displayed

  • CD/Floppy Drives Accept Illegal Character Input

    Description: If use a Web Access client to configure a virtual machine, and you choose Configure Floppy or Configure CD/DVD, files/paths with invalid characters such as ? are accepted.

  • CD/DVD and Floppy Drives Accept Illegal Character Input

    Description: If you use VI Web Access to configure a virtual machine, and you choose Configure Floppy or Configure CD/DVD, files/paths with special characters such as ? are accepted.

  • Auto-Detect Option Not Shown Checked on Newly Added CD-ROM

    Description: The auto-detect option is not shown checked on a newly added CD-ROM.

  • Auto-Detect Option Not Shown Checked on Newly Added CD-ROM for a Virtual Machine

    Description: The auto-detect option is not shown checked on a newly added CD-ROM.

  • S/W iSCSI: Core Dump Partitions Description: Core dump partitions should not be placed on iSCSI LUNs presented over a software initiator.

  • Spanning Extents is Limited to 32 Extents Description: Spanning to a 33rd extent appears to succeed, but does not actually succeed.

  • Qlogic Driver might Freeze

    Description: The Qlogic driver might be unresponsive for 30-60 seconds after one of these messages: "Port database changed" or "RSCN database changed -0x1,0xb00". These messages are typically the result of a cable pull, a host reboot, a SAN reconfiguration change, a flaky HBA, etc., somewhere on the SAN. This happens only with the Qlogic driver.

  • Purple Screen Caused by Lost Connection With SAN (Rare)

    Description: If you trigger certain filesystem metadata operations (open, close, extend, etc.) or reboot, and connection with the SAN is lost at that precise time, e.g. because of SAN failure of LUN availability, a purple screen might result.

  • No Error If Creating More Than 32 Partitions

    Description: The maximum number of partitions with ESX Server is 32. Currently, you can create more than 32 partitions using vmkfstools and the system fails silently (no error, no partition is created).

  • File Operations Between the Service Console and iSCSI or Fibre Channel VMFS Volumes might Terminate Upon Failover

    Description: If you execute long duration file operations such as importing files using vmkfstools or copying VMFS volumes that reside on iSCSI or Fibre Channel SANs, file operations between the service console and VMFS volume might terminate during HBA or SP failover events.

  • Cannot Run ServeRAID 8i and SAS (adp9410) Controllers in the Same System

    Description: Having both a ServeRAID 8i and a SAS (adp9410) controller in the same ESX Server system is not currently supported.

  • Template Upgrade Fails on ESX Server 2.0.1 Host

    Description: When upgrading from VirtualCenter 1.x to 2.0, and managing an ESX Server 2.0.1 host, upgrading of templates fails with the following error message: Failed to communicate with the remote host, either due to network errors or because the host is not responding.

  • VI Client Doesn't Display NFS-Based Datastores

    Description: When ESX Server boots, it remounts all previously configured NFS-backed datastores. If for any reason the first mount request fails, ESX Server keeps retrying until the mount succeeds. When the mount succeeds, he datastore should change from being inaccessible to accessible. However, the VI Client continues displaying the datastore as inaccessible.

  • Changing the IP Address of the VMkernel TCP/IP Stack Results in Temporary Loss of NAS Datastores

    Description: If you change the IP address of a VMkernel TCP/IP, your NAS datastores become unavailable for approximately one minute, and are then returned to service.

  • Special Characters Appear Incorrectly in User-Defined VI Client Fields

    Description: If you use special characters when specifying virtual machine names or other information using the VI Client, they might appear incorrectly.

  • Special Characters in Virtual Machine Names Can Crash VirtualCenter Server

    Description: Special characters can cause VirtualCenter Server to crash.

  • VI Client Login to a Host Fails if the User Doesn't Have Appropriate Permissions

    Description: If a user who doesn't have permissions on the host inventory object tries to log into the host using VI Client, the login attempt fails even if they have permissions on one or more virtual machines or resource pools. This problem manifests in one of two ways:

    • The VI Client hangs when attempting to load the main page.
    • The VI Client displays an unhandled exception message.

  • Virtual Machine with Two or More Virtual CPUs Does Not Power On

    Description: A Virtual SMP add-on license must be configured before virtual machines with more than one virtual CPU can power on. If the Virtual SMP license is enabled on an ESX Server host after the host has been added to VirtualCenter, virtual machines that have more than one virtual CPU might fail to power on with an error message that states: "Virtual machines has x CPUs, but the host software only supports 1."

  • Virtual Machines Fail to Power On After a Snapshot Operation

    Description: Virtual disks added to a virtual machine that aren't in the same directory as the configuration file will cause snapshots to fail.

  • Hot-Adding Disks to Different Datastore (Volume) Than the Base Virtual Machine and Using Snapshot Operations on These Virtual Machine Might Result in Snapshot Operation Failures, Virtual Machine Assert

    Case 1
    Description: If the disks added and the base disks of the virtual machine reside on different datastores, then the virtual machines maintain the same file name, resulting in file search path confusion. The end result of this might be that the virtual machine refuses to power on, virtual machine asserts, or missing snapshot data if you perform snapshot operations.

    Case 2
    Description: Mixing RDM modes (physical and virtual) on the virtual machines with snapshot operations might result in I/O errors inside guest virtual machines, virtual machines boot sector failures, and so on.

    Case 3
    Description: Repetitive combinations of snapshot operations, such as create, revert to, delete, suspend/resume, are combined with adding or removing disks along with different disk modes might cause odd behaviors; for example, disks show removed from the UI but still exist when accessed from the virtual machine, some journal commit I/O errors inside guest operating systems, and, on few occasions, vmware-hostdservicee dies.

  • Powering on the Virtual Machine Results in "File Missing" Pop-up Error Message During Interactions Between Hot-Adding, Adding, and Removing Disks and Snapshots

    Description: The removed virtual disk entries remain in the .vmx configuration file during different cases involving multiple operation of adding disks and snapshotrevert too and go to operations:

    1. Take a snapshot on the virtual machine.
    2. Add a disk.
    3. Revert to snapshot.
    4. Reconfigure the virtual machine in any manner, such as change any setting on the virtual machine.
    5. Power on the virtual machine.

    The virtual machine fails to start.

  • Running Version 2 Virtual Machines that Have RDMs in Append Mode or Undoable Mode Is Problematic

    Description: If you attempt to power on a virtual machine with nonpassthrough RDM append and undoable modes after an ESX Server 2.x to 3.0 upgrade and VMFS-2 to VMFS-3 upgrade, the guest operating system, whether Linux or Windows, panics with the message "The file is missing or corrupt: \WINNT\SYSTEM32\CONFIG\SYSTEMced. You can attempt to repair this by starting the windows installation and selecting "r" in the first boot screen to start repair.

  • Virtual Machine Power On Fails With Error Message "Virtual disk 'Hard Disk 1' is not accessible on the host: Unable to access file '[]'"

    Description: A virtual machine that has an RDM and resides on an ESX Server 2.5 host can't be powered on using VirtualCenter.

  • Migration with VMotion Fails with Error "General Fault Caused by File"

    Description: This occurs when a virtual machine's .vmx file contains the entry: "scsi0:0.deviceType = "scsi-cdrom"" Occasionally, this type of entry is incorrectly added to the .vmx file when the properties of virtual machine are edited from the MUI.

  • SCSI Controller Type Cannot Be Changed in virtual machineSettings If the Virtual Machine Has More Than Three SCSI Controllers

    Description: In a virtual machine with more than three SCSI controllers, if you attempt to change the controller type in virtual machineSettings, the operation fails with the error message: vim.fault.OutofBounds--A parameter exceeded the acceptable range of values.

  • Error After Implicitly Adding Two SCSI Controllers in One Session

    Description: When clicking OK after implicitly adding two SCSI controllers in one session, the error message "the device '2' being referred to for edit/remove device operations cannot be found" appears.

  • DRS: VirtualCenter Failure If Removing Host And Resource Pool Concurrently

    Description: If you are using a DRS cluster, and remove a host and a cluster at the same time, VirtualCenter exits. In addition, the SQL Server gets into a deadlock and rolls back one transaction.

  • Solaris 10 Update 1 Virtual Machine Freezes When a Remote CD-ROM Backend Is Connected in Raw Mode through VirtualCenter Client

    Description: When you connect a remote CD-ROM backend in raw mode through a VirtualCenter Client, the virtual machine freezes and the Windows host machine's CD-ROM drive might become locked and unusable.

  • Uninstalling VMware Tools for Solaris 10 Causes X Server Display System to Be Unusable

    Description: Uninstalling VMware Tools for Solaris 10 causes the X Server display system to be unusable. When the X server is restarted, no graphics are displayed.

  • Red Hat Enterprise Linux Advanced Server 2.1 Virtual Machine Installed with the lilo Boot Loader Does Not Boot Properly after Installation of Update 7

    Description: This problem is caused by a bug in the Red Hat Enterprise Linux Advanced Server 2.1 Update 7 RPM. When kernel-2.4.9-e.62.i686.rpm is installed, the /boot/grub/grub.conf file is updated to refer to the newly built .62 versions of the vmlinix and initrd in the /boot directory. However, the /etc/lilo.conf file still refers to the original .57 versions from U6. When the system is rebooted, the lilo boot loader is not able to find the correct linux image. This in turn causes an invalid SCSI command to be sent to the lsilogic emulation layer.

  • Trying to Eject the Client CD-ROM Device from the Guest Operating System Fails a nd Locks the Client's Drive

    Description: When this problem occurs, if you try to eject a CD/DVD-ROM using the eject command in Linux virtual machines, the operation fails with the error:
    The cd-rom drive failed to eject with error: Incorrect Function. Please make sure the CD-ROM is not in use by the host or attached to any other virtual machines.

    In this case, the CD or DVD cannot be removed from the CD/DVD-ROM drive, and disconnecting the Client drive from the virtual machine does not help. At this point, your CD/DVD-ROM drive is locked and all other virtual machines, including Windows virtual machines, will exhibit the same problem. Virtual machines can browse the contents of the CD/DVD-ROM drive, but you cannot remove the CD or DVD from the drive, even by physically pressing the eject button of the CD/DVD-ROM drive.

  • "Critical vmkernel component(s) failed to load" Error Message Appears After You Log Into Service Console

    Description: After logging into the ESX Server service console, the following error message might be displayed in the service console: "Critical vmkernel component(s) failed to load. Service Console is running in fallback mode. Please file a problem report including output of vm-support."

    If you get this error, the ESX Server system is not functional and you cannot run virtual machines. There are several possible causes of the problem, which is why the message suggests you file a problem report. Possible causes include any of the following:
    This might be a problem with your memory configuration. If, after running "dmesg" from the Service Console, you see two these two messages in your log file, then you have a memory configuration problem:

    "No pages allocated to Node 1 -- big mismatch between BIOS and SRAT memory maps, or MTRR error, or user removed all memory from a Node. Try checking memory or upgrading BIOS"

    "Memory manager could not start (bad0013)"


  • Unable to Power Off Virtual Machine through VI Client

    Description: In ESX Server 3.0, the toolbar power-off button performs a soft power off by default. The soft power off requires the an up-to-date version of VMware Tools to be installed in the virtual machine. If VMware Tools is not installed, or if a legacy version of VMware Tools is installed, the soft power off will have no effect.

  • Browsing for ISO or Floppy Images When Editing the Settings for a Virtual Machine Will Not Show the Service Console File System

    Description: This functionality is currently being worked on and is not available in the current beta release, but it will be available in the final product.

  • After Successful VirtualCenter Database Upgrade, Alarms and Performance Data Is Not Displayed in the VI Client

    Description: After a successful VirtualCenter database upgrade, Alarms and Performance data is not displayed in the VI Client. Although the data is successfully migrated to the VirtualCenter 2.0 database, it is not displayed.

    Workaround: None.

  • Intermittent SSL Warnings Appear During Logout

  • Accessing VI Web Access Through Microsoft Internet Explorer might Cause VirtualCenter to Fail

    Description: When using the VI Web Access proxy to access VirtualCenter (http://{VC_Server}/ui), clicking on Internet Explorer's Refresh button might cause VirtualCenter to fail.

  • Using Drag-and-Drop to Initiate Migrations Incorrectly Skips Host Compatibility Checks

    Description: If you use a drag-and-drop technique to migrate a virtual machine to a standalone host, the Select Host step of the Migration Wizard is skipped. As a result, any pre-migration compatibility warnings/errors are not displayed, and you are incorrectly allowed to complete the wizard and initiate the migration. If there were any compatibility errors, the migration fails after it is initiated.

  • Automatic Update of Virtual Machine VMware Tools

    Description: esx30/VC20 now supports the ability to install or upgrade VMware tools on multiple virtual machines at a time, without needing to interact with each virtual machine. Detailed instructions are provided in the Installation and Upgrade Guide."