6.6.1 Release Update

  • Modified Date:
  • Article ID:000008310


Release Update NB_PDE_6.6.1.tar provides cumulative fixes to Symantec Veritas NetBackup (tm) PureDisk 6.6.


Name: NB_PDE_6.6.1
Date: August 24, 2010

This Patch provides features and fixes to the Symantec NetBackup
PureDisk 6.6 software.










This Patch can be installed on PureDisk 6.6, PureDisk,
PureDisk, and PureDisk

For detailed information about the PureDisk Hotfix, the PureDisk Hotfix, or the PureDisk Hotfix, refer to the Related
Documents section at the bottom of this article, or visit the following



PureDisk automatically creates upgrade jobs for the agent software on
backup and restore clients. The upgrade does not create agent upgrade
jobs for PDDO agents on NetBackup media server clients. To upgrade media
server clients, follow the procedure "Upgrading PDDO client agents on
media servers" in this file.


The following features are new to PureDisk with this release:

* 32 TB support for content routers

* Platform proliferations for the following:

  - AIX 64-bit PDDO support

  - Solaris 10 x64 Remote Office Agent

  - Mac OS 10.5 agent support

  - SUSE Linux Enterprise Server 11 Remote Office Agent

* Rollback for upgrades and patches

* Disaster recovery enhancements

* Internationalization support for the following PureDisk components:

  - Exchange granular restore agents

  - PureDisk core

  - Oracle agents

* Implementation of LZO compression for PDDO

* Improved content router performance

* Multi-NIC support

* OST plug-in for IBM AIX

* Optimized deduplication for the following:

  - VMware Flashbackup VCB for PDDO

  - NetBackup Hyper-V Backups

  - NDMP to PDDO disk backup

* PDOS renamed to PDLinux

* Time zone reset for nodes on a storage pool

* Response time administrative web UI enhancements

* Solaris 10 x64 OST Plug-in

* Graphical representation of storage capacity on the PureDisk Home page

* Upgrade to Storage Foundation 5.0 MP4

* Windows 2008 R2 Support with BEDS SDK


This release contains the following product fixes:

* Etrack 1363394: VSS Snapshot for SQL/Exchange backups fail when
  Backup Exec System Recovery 8 and PureDisk are installed on the same
  Agent host

  To resolve this issue, this patch includes fixes to the methods
  PureDisk uses to get and release snapshots of SQL and Exchange
  databases. The improved methods allow successful PureDisk backups on
  hosts that run both PureDisk and Backup Exec System Recovery 8.

* Etrack 1472868: Updating job step status failing with "Authentication
  for user '%s' failed"

  When a job required a job step to run on the Storage Pool Authority
  (SPA) (such as MB-IMPORT), the job step script ran successfully, but it
  was not able to connect to the workflow engine via the web service to
  update the job step status. Instead an authentication failure occurred.
  The authentication issue that caused this issue has been resolved in
  PureDisk 6.6.1 by adjusting and improving workflow processing.

* Etrack 1851848: Need facility to deduplicate client files based on
  modified time, not archive bit

  PureDisk was unable to back up files only on the basis of modification
  time. Improved backup parameters in PureDisk 6.6.1 now let you ignore
  the archive bit during a scan. The result is that you now have the
  option to select only the new and modified files for an incremental
  backup. See "Specifying specific backup features" in the PureDisk
  Backup Operator Guide for more information about these parameters.

* Etrack 1863894: PureDisk 6.6 FA install - Installer freezes when it
  gets to the disk setup window

  This patch corrects an error that prevented the installer GUI from
  discovering disk groups on certain hardware configurations.

* Etrack 1864302: Requesting non-event based SNMP trap job failure
  notification be enabled for PD

  This patch includes the ability to enable non-event based SNMP trap job
  notification. See "About non-event-based SNMP traps" in the PureDisk
  Backup Operator Guide for more information.

* Etrack 1873696: Content router's spooltasks table should reload when
  queue processing starts

  The spooltasks table was not being reloaded when queue processing
  started. This step ensures that the space used by the content router on
  a multi-node environment is freed and queue processing is not slowed.
  Improvements included in this patch to queue processing and rerouting
  have resolved this issue.

* Etrack 1883307: Replication performance improvements

  The length of time for large replication jobs was increasing during the
  second run of the jobs. Improvements to replication job processing in
  PureDisk 6.6.1 have resolved this issue.

* Etrack 1896573: Race condition results in symbolic links linking to
  the wrong path upon restore

  When large numbers of symbolic links were processed by the backup, they
  may have incorrect targets on restore. Performance and handling issues
  in regard to symbolic links, particularly when backed up or restored in
  high volume, have been addressed in this release. For more information
  about this issue, see Technote 337544

* Etrack 1899417: PDDO 6.6 and VMware VCB backups fail with status 84 -
  media write error

  Etrack: 2070176: NBU 6.5.4 and VCB backups to PDDO fail with status 84

  PDDO 6.6 and NetBackup 6.5.4 backups to a PDDO storage unit failed with
  a status code 84. The bptm log contains the error "fbu_fill_bitmap
  ERROR". Modifications to PDDO have resolved these issues.

* Etrack 1911851: Not all drives that were backed up appear on the
  PureDisk GUI

  Some drives that were successfully backed up were not visible in the
  PureDisk GUI, even though the same drives could be successfully
  restored using the command line interface. This patch corrects the
  error that caused this issue.

* Etrack 1925075: Rerouting running at only 15 MB per second and
  causing high I/O wait states on storage pools

  This patch corrects rerouting progress information in storaged.log and
  includes improvements to rerouting processes.

* Etrack 1941032: Rerouting should not start/resume when the content
  router cache load is not complete

  Rerouting was starting before the cache reload completed. This resulted
  in slow rerouting and cache reload due to the high number of hits and
  queries to the content router database. This patch contains a fix that
  allows the cache reload to complete successfully before rerouting

* Etrack 1941898: Performance degradation due to static content router
  cache settings

  With 240 million objects, the number of entries per hash bucket
  increased up to 80. This slowed content router performance with a
  negative effect on queue processing and rerouting. With this patch,
  PureDisk cache automatically adjusts the hash table size so that the
  maximum list length is limited to 10 entries. Additional fixes to
  adjust yield and allocation unit size provide for quicker loads.

* Etrack 1945690: Minimum operation window for rerouting - rerouting
  should be able to run for a defined window without being interrupted by
  queue processing

  Rerouting now has a defined window in which it runs without being
  interrupted by queue processing. Previously, rerouting was halted when
  queue processing began. If queue processing took more than 12 hours
  rerouting was never able to run. Rerouting now has a one hour window in
  which it is allowed to run.

* Etrack 1952867: Need dynamic heap size settings for Tomcat servers

  Changes to how PureDisk handles the maximum heap setting were included
  in PureDisk 6.6.1. The max heap value was originally set to 512 MB.
  This was insufficient for some systems. The changes allow PureDisk to
  set the max heap value to the lesser of 1 GB or 25% of the available

* Etrack 1959342: Agent does not retry on DNS resolution failure

  PureDisk 6.6.1 incorporates changes to how the PureDisk agent handles
  name resolution. Previous versions of PureDisk made only two retry
  attempts after a failed name lookup. The agent now makes five retry
  attempts when performing name resolution.

* Etrack 1975187: Install GUI incorrectly detects disks.

  Changes to PureDisk allow the installation UI to correctly determine
  the disk quantity and disk path when Dynamic Multipathing (DMP) is
  implemented. With DMP, a disk may have more than one pathname in the
  /dev/ directory, but there is only one entry in the /dev/vx/dmp/
  directory. The PureDisk installation now correctly reports the disk
  quantity and disk path in the computer.

* Etrack 1975354: NBU exports fail after running recoverCR due to
  corrupt SDOREF segments

  Updates to PureDisk let you ignore error messages when exporting data
  selections to NetBackup. Previously, if you recovered a content router
  with the recoverCR tool, you might not be able to export some data
  selections to NetBackup. This can lead to the storage pool becoming
  full, since NetBackup cannot expire the information in PureDisk.
  Changes to PureDisk let you ignore errors in the export process which
  allows NetBackup to expire the data.

* Etrack 1982230: Large numbers of agents (198) fail during overnight

  Older versions of PureDisk could see failed backups when agents
  generated very large (in excess of 10 GB) path object lists. Updates to
  PureDisk 6.6.1 have resolved these issues.

* Etrack 1994888: MonitorCR_ReRoutingMonitorCR fails whole job if one
  connection attempt fails

  Rerouting jobs failed after 1 1/2 days due to one web service
  connection timeout while collecting job progress information. With this
  patch, the number of retries to collect progress information has been
  increased to resolve this issue.

* Etrack 1998466: AIX agent backups hang and eventually time out

  Previous versions of PureDisk had intermittent issues with AIX agent
  backups. The backup would appear to stop responding and eventually time
  out. Changes were incorporated into PureDisk 6.6.1 to resolve these
  timeout issues.

* Etrack 2006852: PDagent installer overwrites pd.config with default

  In previous versions of PureDisk, installing the PureDisk agent on a
  NetBackup media server would overwrite the existing pd.conf file. If
  there were customer-set values, these values were lost. This issue is
  fixed in PureDisk 6.6.1.

* Etrack: 2016331: Request to include bonnie++ in PDLinux

  Bonnie++ is packaged with PDLinux starting with PureDisk 6.6.1.
  Bonnie++ lets you measure disk performance.

* Etrack 2026043: PDRAI is faulting intermittently on some file system

  In some cases, the pdrai.exe application generated errors
  intermittently during file system backups. Code corrections to the
  application that resolve the problem are included with PureDisk 6.6.1.

* Etrack 2030549: Updating configuration parameters in GUI results in
  status 500 error

  When attempting to save a configuration value in the GUI, the operation
  returned a status 500 error ("The server encountered an internal error
  () that prevented it from fulfilling this request."). PureDisk 6.6.1
  includes a fix so that configuration changes are successfully saved and
  the status 500 errors no longer occur.

* Etrack 2049216: PureDisk data mining report shows 1/4 of the
  total numbers of data selections

  In PureDisk, the data mining report did not show all the data
  selections for a VCS clustered storage pool. This issue is resolved in
  PureDisk 6.6.1.

* Etrack 2051628: Unable to browse directories with a comma in the name

  Users were unable to browse directory names that contained a comma from
  the Restore Files interface. With PureDisk 6.6.1, the entire directory
  can be restored successfully and can also be successfully restored
  using the Find Files interface.


The following are the known issues associated with this patch:

* Etrack 2139357 - After applying the PD661 patch creating a DS list or
  restore to alternate location fails once changed configuration to
  single port

  This issue occurs if you switch to a single port configuration after
  you apply the PureDisk 6.6.1 upgrade patch. In this case, you are no
  longer able to browse client resources when trying to create data
  selection lists or when trying to restore to alternate locations. If you
  create a data selection list and try to select resources for backup,
  you may receive the following error message:

   "unable to execute dirlist on agent".

  Workaround: To avoid this issue, do not switch to single port
  communication after upgrading to PureDisk 6.6.1.

* Etrack 2110667: Disk pool goes down and content router is
  unresponsive due to web service error

  NetBackup-optimized duplication jobs to PureDisk storage pools may
  timeout if there are too many concurrent jobs.

  Workaround: To resolve this issue, take one of the following actions:

  - Reduce the number of concurrent duplication jobs to match the
    concurrency level in the PureDisk storage pool by restricting the
    concurrency in the NetBackup storage unit.


  - Edit the /etc/puredisk/pdwfe.cfg file, and increase the
    "taskthreads =" value (that is, the number of threads in the workflow
    engine) from 4 to a larger number for both source and destination
    storage pools. Then restart the services.

* Etrack 2106313 - When VCS number and cluster ID are same, the cluster
  upgrade install fails

  When the VCS number and the cluster ID are the same, the cluster
  upgrade fails with the following error message:

  PERSISTENT_NAME is not set for all the nic's you need to set them manually before the next reboot.

  Workaround: If this issue occurs, change the VCS number or the cluster
  ID so that they are not the same, and then rerun the cluster upgrade.

* Etrack 2084452 - One of the nodes in the cluster failed to come up on
  reboot after kernel upgrade

  After the reboots of any set of nodes, the following message may appear
  in the VCS engine log:

   2009/02/06 11:42:35 VCS CRITICAL V-16-1-11306 Did not
   receive cluster membership, manual intervention may be
   needed for seeding

  Workaround: If the message appears, restart VCS by performing the
  following steps:

   1. Kill had and hashadow:
   # pkill had; pkill hashadow
   2. Wait for 10 seconds and then run the following command:
   # hastart

  VCS should restart without any issues, and the earlier message should
  not reappear in the VCS engine log.

  More detailed information is available in Veritas Storage Foundation
  and High Availability Solutions - Read This First for Linux - 5.0
  Maintenance Pack 3 Rolling Patch 2.

* Etrack 1989561 - After reboot in Directory Services Restore Mode, PD
  Agent service is not starting on Windows 2008 R2 agent with active
  directory installed on it.

  Because the DNS client service does not start in Directory Services
  Restore Mode, the name resolution does not work properly. Service
  Symantec NetBackup PureDisk Client Agent does not start automatically
  when the agent machine is restarted in Directory Services Restore Mode
  on Windows 2003 R2 or on Windows 2008 R2. This occurs for an agent
  upgraded from PureDisk 6.6 or PureDisk to PureDisk 6.6.1 and
  also for a newly added agent on PureDisk 6.6.1.

  Workaround: Add the IP address of the SPA server to
  c:\WINDOWS\system32\drivers\etc\hosts. To do this, open the hosts file
  in a text editor and follow the instructions in the file from Microsoft
  for adding entries.

* Etrack 2085387 - After applying the PureDisk 6.6.1 patch in an iSCSI
  environment, existing disk groups still show as 'LVM'

  PureDisk 6.6 includes Storage Foundation 4.x which uses LVM for
  management of iSCSI environments. The version of VxVM in Storage
  Foundation 4.x does not support iSCSI. PureDisk 6.6.1 includes Storage
  Foundation 5.0 MP4 which allows you to use VxVM for management of iSCSI
  environments. After an upgrade to PureDisk 6.6.1, existing iSCSI disks
  continue to be managed using LVM. The disk groups are not converted to
  VxVM. Only new disk groups created after the upgrade to PureDisk 6.6.1
  use VxVM for management of iSCSI disks.

  Workaround: To use VxVM instead of LVM to manage iSCSI devices, you
  must create new disk groups after you successfully upgrade to PureDisk

* Etrack 2085373 - After applying the PureDisk 6.6.1 patch in an iSCSI
  environment, the mount point for a second node is cleared.

  This issue applies to situations where an iSCSI device is connected to
  two nodes, and the two nodes share Logical Volume Manager (LVM) group
  information. Even if upgraded to PureDisk 6.6.1, existing LVM-created
  disks and groups are still managed by LVM.

  When the PureDisk 6.6.1 patch is applied, the mount point for the
  second node does not appear because LVM does not have a function to
  distinguish the machine on which you have created groups.

  Workaround: Symantec recommends that you do not connect the same iSCSI
  device to multiple nodes that share LVM groups. If you still want to
  connect multiple nodes to an iSCSI device, do not map the same iSCSI
  logical unit numbers (LUNs) to different nodes. (Run the "ls -l
  /dev/disk/by-id/" command on each node to make sure that there are no
  matching IDs on the different nodes.)

* Etrack 2085391 - After applying the PureDisk 6.6.1 patch in an iSCSI
  environment, used LUNs appear as free for a newly added node.

  This issue occurs using the configuration wizard in an iSCSI
  environment when you create a second node that is connected to an iSCSI
  device and then try to create a disk group on the second node. Logical
  unit numbers (LUNs) that were previously assigned to another node
  appear as free. If you try to add these LUNs to a new disk group on the
  second node, an exception message similar to the following example

   initialize disk failed with: Can't initialize physical volume "/dev/sdb" of
   volume group "PDDG14" without -ff

  Workaround: Symantec recommends that you do not connect the same iSCSI
  device to multiple nodes that share disk groups. If you still want to
  connect multiple nodes to an iSCSI device, do not map the same iSCSI
  logical unit numbers (LUNs) to different nodes. (Run the "ls -l
  /dev/disk/by-id/" command on each node to make sure that there are no
  matching IDs on the different nodes.)

* Etrack 2085378 - After applying the PureDisk 6.6.1 patch in an iSCSI
  environment, the disk groups are visible on all nodes.

  In an iSCSI environment, the disk groups created on one node are
  visible on all nodes in the topology.

  Workaround: Symantec recommends that you do not connect the same iSCSI
  device to multiple nodes that share disk groups. If you still want to
  connect multiple nodes to an iSCSI device, do not map the same iSCSI
  logical unit numbers (LUNs) to different nodes. (Run the "ls -l
  /dev/disk/by-id/" command on each node to make sure that there are no
  matching IDs on the different nodes.)

* Etrack 2104003 - Searching for emails with "- >>" in subject returns
  no results

  When searching for emails with a Subject field that contains the
  characters "- >>" (including spaces), the search returns zero (0)
  results due to an issue with Microsoft Exchange.

  Workaround: If you need to find email items with "- >>" in the Subject
  field, you must page through all emails in the Inbox and select items
  individually for restore.

* Etrack 2101661 - On a Simplified Chinese Exchange 2003 setup, the
  Exchange database is not mounted automatically after a point-in-time
  restore with the "mount after restore" option

  After a successful point-in-time restore of an Exchange 2003 data
  selection in a Simplified Chinese operating system environment, the
  Exchange 2003 database is not mounted automatically even if the "mount
  after restore" option is specified. No error or warning messages are
  issued in this situation. The restore job otherwise completes
  successfully, and the data is restored.


  This issue applies only to Exchange 2003 setups that use Simplified

  Workaround: If you use a Simplified Chinese Exchange 2003 environment,
  you must manually mount the Exchange database after a successful
  restore, even if you specified the "mount after restore" option.

* Etrack 2035775 - PureDisk does not detect a failure to mount an
  Exchange 2007 database in some restores

  For some Exchange 2007 database restores, you may experience a
  situation in which the restore job finishes with a status 0 (success)
  and with "Mount database after restore" indicated as successful.
  However, the database fails to restore the Exchange 2007 database to a
  mountable and usable state.

  Workaround: If you encounter this issue, you must manually mount the
  Exchange 2007 database after the restore. If the manual mount fails,
  attempt to rerun the restore job as a point-in-time Exchange restore to
  correctly restore the data.

* Etrack 2005904 - Items that exist in the Deleted Items box prevent a
  Exchange granular restore from writing those items to another folder in
  the mailbox

  The issue occurs only if the original backed up folder was moved to
  Deleted Items and a restore of the deleted folder is restored back to
  its original location. If the restored folder is placed in Deleted
  Items, the restore works as expected. For example:

   1) Back up a mailbox that has Folder-1 with five items.
   2) Delete Folder-1 and remove it from Deleted Items.
   3) Restore Folder-1. Restore works.
   4) Delete Folder-1 and leave it in Deleted Items.
   5) Restore Folder-1. Restore also works.

  The problem occurs only if the original backed up folder is placed in
  Deleted Items. If you perform a restore of that folder, the items in
  the folder are rerouted to Deleted Items by MAPI.

  Workaround: The Deleted Items folder should be emptied before you
  perform a restore. In most cases, this is not necessary because you
  typically restore granular data when they are lost or already deleted.
  If you want to restore items while they are still in Deleted Items, you
  must empty Deleted Items before you perform the restore.

* Etrack 1998530 - PureDisk 6.6.1 is restoring extraneous log files
  with certain Exchange storage group restores, causing the database to
  fail to mount

  This issue is the result of a problem with Microsoft Visual SourceSafe
  (VSS). On a restore of a storage group with a public folder database,
  the VSS Exchange writer is returning many log files that do not need to
  be backed up. The log files are restored as 0-length files, and they do
  not appear in the catalog list of files. They cause the mount of the
  Exchange database to fail because they are empty.

  Workaround: To resolve this problem, manually delete the 0-length log
  files that were restored and then manually mount the Exchange database.

* Etrack 2088563 - Search function is not working for Public folder in
  Exchange Restore Files window

  Items in Public folders that match search parameters are not displayed
  as search results in the File Selection page of the Exchange Restore
  Files wizard. The item selection box of the matching items is removed,
  and the wizard does not let you select the items for restore.

  Workaround: If you need to search multiple items with the same subject
  or a similar subject, Symantec recommends that you search and restore
  the items for one subject at a time. Avoid using the "Clear last
  search" option to perform a consecutive search on the same subject or
  similar subjects. However, you may use "Clear last search" in between
  consecutive searches if the subject to be searched is different.

  To search for an item, enter search text in the search box and click
  GO. Symantec recommends that you use GO instead of pressing ENTER to

* Etrack 2048171 - The granular restore message is "Symantec Backup
  Exec" instead of "Symantec PureDisk"

  When PureDisk completes a Microsoft Exchange granular restore, the
  restored mailbox receives an email message. The subject reads "Symantec
  Backup Exec for Windows Servers initialization message for mailbox
  restore" instead of "Symantec NetBackup PureDisk initialization message
  for mailbox restore." The text of the message also mentions Symantec
  Backup Exec instead of Symantec NetBackup PureDisk.

  Workaround: Please ignore the mention of Symantec Backup Exec in the
  Subject and text of the initialization message.

* Etrack 1826217 - Mac OS X agent - backups failing after longer time
  usage on DNS hostname resolution

  In certain environments, the Mac OS X client is unable to resolve
  hostnames even though it can connect to the hosts with IP addresses.
  Backups fail with a hostname resolution failure. After the pdagent
  restarts, the client can resolve the hostname and the backup succeeds.
  This issue reoccurs at frequent intervals making daily backups almost

  This problem is a known issue for Mac OS X. More information about this
  issue is available at the following site:

  Workaround: For backups of Mac OS X agents, specify IP addresses
  instead of the DSN hostnames.

* Etrack 2087799 - High percentage of PDDO backups fail with status 84
  when content router rerouting is running

  When content router rerouting is running, a high percentage of PDDO
  backups fail with status code of 84.

  Workaround: If you experience this issue, disable caching by adding
  "CACHE_DISABLED = 1" to the pd.conf files on all media servers.

* Etrack 2046312 - After the upgrade, first new job will have more data
  for PDDO

  When you run the first backup job after you upgrade to PureDisk 6.6.1,
  you may observe a lower deduplication rate than you observed before the
  upgrade. This difference is due to changes in the backup mechanism for
  the PureDisk Deduplication Option (PDDO). Because less data is
  deduplicated, PureDisk backs up more data. Consequently, there is the
  potential of overloading storage pool resources.

  Workaround: Ensure that you have adequate capacity in the storage pools
  for large amounts of backup data.

* Etrack 2073490 - VSS does not work for symbolic links and junction
  points on same and different drives

  Backup jobs fail if they are created with Microsoft Visual SourceSafe
  (VSS) (Use snapshotting) with the option "follow symbolic links" or the
  option "follow junction points" on the same drive or on different

  Workaround: Do not create backup jobs using VSS with the "follow
  symbolic links" or the "follow junction points" options.

* Etrack 2079582 - Appropriate error message is not displayed when the
  AD user selects a department for which the user is not authorized

  A user with only View permission on an agent should not be authorized
  to access the department to which the agent belongs. If that user tries
  to access the department level, the following message appears:


  Workaround: The message indicates that the user has not been authorized
  to access the department level. Users who receive this message should
  contact their administrator or try instead to access the agent level
  for which they have been authorized.

* Etrack 2034544 - Individual files cannot be selected for restore from
  GUI if the data lock password is set

  Symantec has discovered issues with data lock passwords that contain
  HTML reserved characters. If the data lock password contains HTML
  reserved characters, individual files and folders cannot be selected
  for restore. There is no risk to data, since the entire data selection
  can be restored. See the PureDisk Getting Started Guide for more
  information about the data lock password format.

  Workaround: To prevent this problem, do not use any HTML reserved
  characters in the data lock password. The HTML reserved characters are:

   - Double quotation mark (")
   - Single quotation mark (')
   - Ampersand symbol (&)
   - Less than symbol (<)
   - Greater than symbol (>)

* Etrack 1922902 - Receiving a segmentation fault message at the end of
  a disaster recovery restore of PureDisk 6.6.1

  When performing a disaster recovery restore from an encrypted Samba or
  third-party disaster recovery backup, you may receive the following
  message at the end of the restore process:

   ./DR_Restore_all.sh: line 14: 8318 Segmentation fault     
   $PHPBIN /opt/pdinstall/lib/DRRestoreAll.php

  Workaround: You may safely ignore the message. This message does not
  affect the restore or system functionality.

* Etrack 2036628 - Receiving wrong text message when "Cluster java
  console" link is selected

  Because of a bug in Internet Explorer, you may not be able to download
  items such as "Cluster java console" from the configuration wizard
  login page in an Internet Explorer browser window.

  Workaround: Use Mozilla Firefox instead of Internet Explorer. PureDisk
  also supports Mozilla Firefox.

  To upgrade the Java Console on a Windows client with Mozilla Firefox,
  perform the following steps:

   1. Stop Cluster Manager if it is running.
   2. Remove Cluster Manager from the system.
   3. Using Mozilla Firefox, download the Cluster Manager installer from the
      Storage Pool Configuration Wizard welcome screen. (http://<spa address>/Installer)
   4. Run setup.exe
   5. Follow the wizard instructions to complete the installation.


This table lists the publications and their revision levels for PureDisk
Release 6.6.1.

  | Manual title                                  | Revision level  |
  |  PureDisk Administrator's Guide               |    6.6.1        |
  |  PureDisk Backup Operator's Guide             |    6.6.1        |
  |  PureDisk Best Practices Guide                |    6.6.1        |
  |  PureDisk Client Installation Guide           |    6.6.1        |
  |  PureDisk Deduplication Option Guide          |    6.6.1        |
  |  PureDisk Getting Started Guide               |    6.6.1        |
  |  PureDisk Storage Pool Installation Guide     |    6.6.1        |
  |  PureDisk Command Line Interface Guide        |    6.6.1        |
  |  PureDisk Cluster Planning Spreadsheet        |    6.6          |

The PureDisk 6.6.1 documentation set does not support the NetBackup 5000
appliance at the time of the PureDisk 6.6.1 GA release. PureDisk 6.6.1
support for the NetBackup 5000 is deferred.


This section describes the preinstallation steps for the Patch.

Downloading and extracting the patch

1. Use scp to copy the tar file, NB_PDE_6.6.1_NNNNNN.tar to the /root
   directory of the PureDisk node that hosts the storage pool authority.

2. Log on as root to the node that hosts the storage pool authority.

3. Type the following command to verify the integrity of the

   md5sum /root/NB_PDE_6.6.1_NNNNNN.tar

   This command computes the md5 checksum of the Patch. The md5
   checksum of the Patch must match '99d2c2d02ad315c7541271404d87aac5'.


   If you obtain a different checksum, the Patch was corrupted
   during download. Try to download the Patch again.

4. Type the following command to extract the README file:

   tar -C / -xf /root/NB_PDE_6.6.1_NNNNNN.tar ./NB_PDE_6.6.1.README


This section describes how to install the software.

Installing the patch

1. Make sure that no PureDisk jobs are currently running or are
   scheduled to be run.

2. (Conditional) If you have /Storage partitions that are mounted on
   an NFS share, add an entry for each partition to the /etc/fstab file.

3. Log out from the PureDisk administrative Web UI.

4. (Conditional) Freeze the PureDisk service groups for the clustered
   PureDisk server.

   Perform this step if the storage pool is installed with Veritas
   Cluster Server (VCS) software.

   For each active node, you will have groups such as pd_group1,
   pd_group2, and so on. Log on to the Storage Pool Authority (SPA) node
   and run the following command for each of these groups:

   /opt/VRTS/bin/hagrp -freeze pd_group1

   Where pd_group1 indicates the group you are freezing.

   For information about how to freeze and unfreeze clustered storage
   pools, see the Symantec NetBackup PureDisk Storage Pool Installation

5. Type the following command to unpack the Patch software:

   tar -C / -xf /root/NB_PDE_6.6.1_NNNNNN.tar ./opt

6. Type the following command to run and install the Patch:


   If the topology.ini file is encrypted, the software prompts you for
   the password to decrypt this file.

   The Patch automatically pushes the software to all nodes in
   the storage pool and to all clients.


   The Veritas Cluster Server (VCS) software might detect some faults
   during the upgrade process. If any upgrade actions generate a VCS
   fault, use the Cluster Manager Java Console to clear the fault and
   probe that resource group before you unfreeze the cluster.

   At the end of a successful installation, the software prompts you to
   encrypt the topology.ini file.

   When the upgrade script completes successfully, the script also
   unfreezes the PureDisk service groups for the clustered PureDisk

7. (Conditional) Again, freeze the PureDisk service groups for the
   clustered PureDisk server.

   See Step 4 for instructions about freezing PureDisk service groups.

8. (Conditional) If you previously used Veritas Enterprise
   Administrator (VEA) to manage your storage, download and install the
   new VEA client (Version 3.3.721.0) from the following URL:

   http://<spa address>/Installer

   This step is necessary because PureDisk 6.6.1 is not compatible with
   the previous version of VEA (Version 3.2.540.0), which worked with
   PureDisk 6.6. Upgrading to VEA Version 3.3.721.0 ensures that PureDisk
   6.6.1 can operate properly with VEA.

9. Log on to each node (active and passive) and run the following
   script to install the kernel upgrade:

   /opt/pdinstall/apply-661kernel-patch.sh --upgrade

   If the kernel upgrade fails because some services do not stop, wait
   until the services stop and rerun the kernel upgrade script.

10. After the upgrade jobs finish, reboot all of the nodes.


    Allow the upgrade jobs that were started in step 6 to finish,
    successfully or not, before you reboot the nodes. Symantec suggests
    that you reboot all of the nodes (active and passive) at the same time.


    After the reboot, the clustered service groups that were frozen in
    Step 7 are unfrozen."

11. (Conditional) If you are performing this upgrade as part of
    adding a new node to a cluster, restart VCS by performing the
    following steps:

     1. Kill had and hashadow:
     # pkill had; pkill hashadow
     2. Wait for 10 seconds and then run the following command:
     # hastart

    For information about adding a new node to a cluster, see the
    PureDisk Administrator's Guide.

12. (Conditional) To reconnect with the Veritas Enterprise
    Administrator (VEA) in a clustered environment, run the following

    /etc/init.d/vxpal.StorageAgent start

    /etc/init.d/isisd start

13. Monitor the client agent update jobs

    Take the following actions to monitor the client agent upgrade jobs:

    * Clear the browser's cache and temporary Internet files.

    * Start the administrative Web UI.

    * Click Monitor > Jobs.

    * Set the "View jobs by" dropdown to "Policy types".

    * Select the "Agent Update" workflow under "Miscellaneous

    If one of the upgrade jobs fails, PureDisk deactivates those client
    agents. You need to activate these agents again and to select the
    client to upgrade to. Follow the steps below:

    * Start the administrative Web UI.

    * Click Manage > Agents.

    * In the left pane, select the deactivated agent you want to
      reactivate. Select the storage pool level if you want to activate
      all agents in the storage pool.

    * In the right pane, select "Activate Agent(s)".

    * A new agent update job is scheduled for those agents not
      upgraded yet.

    * Monitor these upgrade jobs again. Start them over if they
      fail or timeout.

14. (Conditional) If your media servers runs NetBackup 6.5.x, upgrade
    the PDDO agent software on PDDO media server clients.

    CLIENTS", which follows. The release update includes upgraded PDDO
    agents for Linux, Solaris, and Windows clients.

    This upgrade is not necessary if your media server runs NetBackup 7.0
    or later


This section describes how to upgrade PDDO agents on media server clients.

Perform this procedure only if your media server runs NetBackup 6.5.x.
This upgrade is not necessary if your media server runs NetBackup 7.0 or


Symantec encourages you to upgrade your PDDO agents at this time. The
PureDisk 6.6.1 release includes enhancements to the compression
algorithm. Performance is degraded when you restore data from a PureDisk
6.6.1 storage pool to a PDDO agent that is older than PureDisk 6.6.1.


The upgrade software removes the old PDDO agent software automatically.
For more information about how to install the PDDO agent, see the
Symantec NetBackup PureDisk Deduplication Option Guide.

Upgrading PDDO

1. Make sure that the storage pool software for the PDDO clients has
   been upgraded.

2. Refresh the PureDisk landing page.

   The landing page can be found on:


   For <your-puredisk-spa>, type the hostname or the IP address of your
   PureDisk storage pool authority.

   Be aware that this Patch includes both the PureDisk 6.6 base
   release agents and revised agents for Linux, Solaris, and Windows
   clients. The revised agents include updates for PDDO agent
   installation only. A later step in this procedure instructs you to
   download the revised agents if you want to upgrade the PDDO agent on a

3. Log on to each PDDO client.

4. From the PDDO media server client, download the new PDDO agent

   The landing page includes more than one software version of the PDDO
   agent, so take care to download the most recent agent.

5. (Conditional) Freeze the PureDisk service groups for the clustered
   PDDO server.

6. Stop all NetBackup services on the NetBackup media server.

7. Install the agent software that is included with this

   * On Windows systems, double-click the PureDisk agent icon. That
     starts the Windows Installation Wizard .

   * On Linux or Solaris systems, you can use either the attended or
     the unattended installation method. The installer prompts you to
     confirm the upgrade. For example, type the following command to
     upgrade the Solaris 10 agent:

   # sh pdagent-Solaris_10_sparc-

8. (Conditional) Unfreeze the PureDisk service groups for the
   clustered PDDO server.

9. Start all the NetBackup services on the NetBackup media server.

Terms of use for this information are found in Legal Notices.



Did this article answer your question or resolve your issue?


Did this article save you the trouble of contacting technical support?


How can we make this article more helpful?

Email Address (Optional)