HP StorageWorks NAS 8000 - Version 1.6.X Instruction Manual

HP StorageWorks NAS 8000 Version 1.6.3 Release
Notes
These release notes are for NAS operating system (OS) version 1.6.3. The version number of the NAS
OS software appears on the Identity tab of the Command View NAS interface.
Figure 1: Identity tab
* A7418-96026*
A7418-96035

These release notes appear in the following main categories:
Product Changes and Improvements (Versions 1.6.2 and 1.6.3)
Release Notes
Operating System and NAS Web Interface
Applications
NAS Data Path Manager
Product Changes and Improvements (Versions 1.6.2 and
1.6.3)
Operating System
High Availability
Modifications to HA failover timeouts to reduce the chance of premature package failovers.
Modifications to HA configuration caching to eliminate cache coherency problems that would
block HA packages from failing over successfully.
Eliminate a problem with the rmtab file in HA systems.
Ensure consistent User Mappings between cluster nodes.
Networking
Updated driver for the NC7770 copper Gigabit Ethernet Network Interface Card (NIC).
Security
Modifications to Samba to close a potential security hole identified by the Open Source
Community.
Snapshots
Eliminate a problem that would occasionally prevent the user from successfully creating a
snapshot.
Eliminate a problem that would occasionally cause snapshots to hang under heavy load.
Eliminate a problem caused when creating a snapshot of the Disaster Recovery Volume.
Backup
NAS 8000 release 1.6.3 contains the new Data Protector 5.1 agent. For more information, see
the Backup section.
Upgrades
Correct several registry problems introduced when upgrading from version 1.4.X to 1.5.0 and
from 1.5.0 to 1.6.X.
Command View NAS Web Interface
Improved GUI performance, particularly when configuring storage.
Supported Hardware
Support for VA7110 in both dedicated storage and SAN configurations.
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 2

Release Notes
Operating System and NAS Web Interface
Upgrading the NAS Server Software
•The procedure for upgrading the NAS server software is presented here. You can also obtain the
most recent upgrade instructions from the support web site at: http://www.hp.com/support/nas.
To upgrade the software:
1. Verify that a good backup of all data has been made.
2. Verify that a recent DRF file is on non-NAS media (floppy or local PC drive).
3. Create a CD containing the following RPM files from the support website:
•hpNasOs-1.6.3-68.i386.rpm
•hpNas-antivirus-1.6-0.i386.rpm
•hpNas-antivirus-authorize-1.6-0.i386.rpm
•If you are using Data Protector 5.1
- hpNas-backupDP-5.1-1.i386.rpm
- hpNas-backupDP-authorize-5.1-1.i386.rpm
•If you are using Data Protector 4.1 or 5.0
- hpNas-backupDP-4.1-1.i386.rpm
- hpNas-backupDP-authorize-4.1-1.i386.rpm
4. Upgrade the NAS OS from the GUI using the hpNasOs-1.6.3-68.i386.rpm file on the CD.
5. Reboot the system.
6. Telnet to the NAS 8000, or use the console to complete the installation.
a. Uninstall Antivirus and the Omniback Agent:
•Type: uninstallSystemPackage hpNas-antivirus-authorize
•Type: uninstallSystemPackage hpNas-antivirus
•Type: uninstallSystemPackage hpNas-backupOB-authorize
•Type: uninstallSystemPackage hpNas-backupOB
•Ignore errors about InvalidSystemArgument (trying to remove files that the upgrade
already removed).
b. Install the new Antivirus:
•Put the CD in the NAS 8000 CD-ROM drive.
•Type: installSystemPackage cdrom://hpNas-antivirus-1.6-0.i386.rpm
•Type: installSystemPackage cdrom://hpNas-antivirus-authorize-1.6-
0.i386.rpm
c. Install the new Data Protector Agent:
•Put the CD in the NAS 8000 CD-ROM drive.
•For the Data Protector 5.1 Agent:
1. Type: installSystemPackage cdrom://hpNas-backupDP-5.1-1.i386.rpm
2. Type: installSystemPackage cdrom://hpNas-backupDP-authorize-5.1-
1.i386.rpm
•For the Data Protector 4.1 Agent:
1. Type: installSystemPackage cdrom://hpNas-backupDP-4.1-1.i386.rpm
2. Type: installSystemPackage cdrom://hpNas-backupDP-authorize-4.1-1.i386.rpm
7. Restart the GUI.
8. Re-add Antivirus scans and updates (be sure to update the signature file immediately).
9. Re-setup Data Protector Agents.
10. Re-import the Data Protector Agent on the Data Protector Cell Manager.
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 3

UPS Settings
•If you connected your HP NAS 8000 to the optional APC Symmetra UPS during installation, the
NAS server attempts to communicate with the UPS through a LAN connection. Please follow the
directions for the Symmetra to configure it first. Once it is configured, you can make the HP NAS
8000 aware of it by entering the information in Command View NAS. See the instructions in the
User’s Guide in the Configuring Your System and Network chapter, Configuring UPS
Connections section.
Command View NAS Tested Configurations
•Any browsers that support the Sun Microsystems Java Plug-in (version 1.3.1 or later) should be
able to run the Command View NAS interface. Tested configurations include:
OS Platform Data Access Browser/GUI Java Plug-in
RedHat 7.3 X NS 4.7x 1.3.1x
RedHat 8.0 X Mozilla 1.3
RedHat 9.0 X Not tested Not tested
Windows 2000 X IE 6.0 1.3.1x
Windows NT 4.0 SP6a X IE 5.5 SP2 1.3.1x
Windows XP X IE 6.0 1.4.2
Windows 98 SE X IE 5.5 SP2 1.3.1x
HP-UX 10.20 X Not supported Not supported
HP-UX 11.0 X NS 7.0 1.4.1x
HP-UX 11i v1 X Not tested Not tested
Sun Solaris 7.0 X NS 7.0 1.4.1x
Sun Solaris 8.0 X Not tested Not tested
Sun Solaris 9.0 X Not tested Not tested
SCO UnixWare 7.1 X Not supported Not supported
SCO UnixWare 7.1.3 X Not supported Not supported
AIX 4.3.3 X Not supported Not supported
Users and Security
•The access control list (ACL) implementation on the NAS 8000 for Windows clients is based on
Windows 2000 semantics, not Windows NT semantics. This may cause confusion if a user is
assuming NT semantics. This primarily affects inheritance behavior. You can install the Microsoft
Windows Security Manager patch on any pre-Windows 2000 systems that will be used to manage
NAS 8000 Windows protocol security.
•Administrators that set up CIFS shares using "share-mode" security may see anomalies with
DOS attributes. DOS attributes, specifically the archive bit, may not exhibit the same behavior as
a native Windows operating system in DOS mode. Network backup packages that rely on the
archive bit for incremental backups may not behave properly (knowing when, or when not, to
backup a file) in share mode security. For example, the DOS archive (A) attribute is not set when
a file is moved from one folder to another.
•A client connected through CIFS on a user-mode security share will be able to execute a file if
they have read permission in the ACL. On Windows operating systems, explicit execute
permission needs to be given.
•When deleting read-only files on a CIFS share, there will be no confirmation dialog box. The file
will just be deleted.
•Window NT clients cannot take ownership of Windows 2000-created objects (files or directories).
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 4

•Do not attempt to join a domain in which the controlling primary domain controller (PDC) does not
have a WINS entry on the WINS server. This may lead to the NAS server not being able to join
any domains in the future, even if the WINS entry is entered later.
•For user mappings to work correctly, the mapping needs to happen prior to the client attaching to
the NAS server. If the client is currently attached and then the mapping takes place, the client
must be rebooted to update the connection information and save any information to the NAS
server with the correct owner.
•After installation, if Windows users are not being correctly mapped, a reboot can fix the situation.
This problem can be identified by checking the mapped users on the system after an attempted
connection. If the user was not mapped, the domain probably did not register correctly, and the
NAS server needs to be rebooted to clear this problem.
•Some Windows clients that have mapped NAS 8000 shares show zero free space available on
shares that in reality have free space. In such a case, there is a textApi command that will direct
the NAS 8000 to use a different way to report free space to Windows clients. Enter the following
command from the NAS 8000 command line:
% setNetworkSmbDfreeCommandEnabled T
Afterward, NAS 8000 shares should correctly report the free space to Windows clients. One
negative side effect is slower CIFS read/write performance. NFS performance is unaffected. To
go back to the default way of reporting free space to Windows clients, enter:
% setNetworkSmbDfreeCommandEnabled F
High Availability Windows Monitor
•In some complex Windows domain environments, the clustered NAS 8000 Samba package
monitor randomly fails to authenticate and thus causes the package to fail over unnecessarily. If
this becomes an intolerable problem, then it is possible to tell the NAS 8000 Samba package
monitor to use a less rigorous monitoring algorithm. The default algorithm is enabled using the
following command:
% setNetworkSmbMonitorCommand "/sbin/smbMonitor"
The less rigorous monitoring algorithm is enabled using the following command:
% setNetworkSmbMonitorCommand "/sbin/smbMonitorLite"
It is possible to create another monitoring algorithm. The argument to the
setNetworkSmbMonitorCommand() command is the absolute pathname of a program or script
that returns 0 when Samba is functioning properly, and returns 1 when it is not. To enable
another algorithm, install the program or script on the NAS 8000, and then enter:
% setNetworkSmbMonitorCommand "<Program>"
where <Program> is the absolute pathname of the new program or script that monitors Samba.
High Availability (HA) Cluster Configuration
•Volume groups that are not a part of a package appear to be available to create file volumes.
However, an error will occur if the user attempts to create a file volume on a volume group that is
not a part of a package.
•Certain storage objects such as file volumes, directories, shares, snapshots, and scheduled
snapshots may be administered, viewed, and managed only via the Command View NAS
interface for the node on which they are active.
•If an HA package is failed over to a node that already has Command View NAS running, it may
be necessary to refresh the Volume Group page before the existing browser session will
recognize the newly failed over storage.
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 5

•On an HA system, users may run into difficulty deleting a volume group if they delete the failover
package containing the volume group without first deleting any file volumes on that volume group.
A volume group cannot be deleted without first removing all file volumes that it contains.
However, once a package has been removed, any associated file volumes will not be
visible/deletable. This may lead to some confusion, since the volume group will appear to be
empty, but the user will get an error message stating that file volumes must be removed before
the volume group can be deleted. A workaround is to create a package that includes the volume
group in question and then delete any file volumes that will now become visible. Then the
package and the volume group may be deleted.
•If the heartbeat path between the two cluster nodes is broken (cable removed, one of the nodes
shut down, etc.), the Command View NAS interface will become unresponsive.
•When configuring a bond in an HA cluster, do not choose individual NICs that are participating in
a bond as heartbeat NICs. Instead, include the bond and/or other network cards. Choosing a
NIC that is participating in a bond will cause the cluster to be misconfigured and could cause
unintentional package failover.
•There is a known issue regarding bonding. When a bond is created, the administrator can add
NICs in any order. However, this order is not preserved when the system boots. The NICs are
added to the bond in alphanumeric order. The practical implications to this are twofold:
•Any browsers that support the Sun Microsystems Java Plug-in (version 1.3.1 or later) should be
able to run the Command View NAS interface. Tested configurations include:
•Any browsers that support the Sun Microsystems Java Plug-in (version 1.3.1 or later) should be
able to run the Command View NAS interface. Tested configurations include:
1. You cannot specify the primary NIC.
- If you bond a 100-Mbit eth0/1 with a 1-Gbit eth2/3, the primary NIC will always be the
lower-numbered (slower) NIC upon booting. This is probably not the intended behavior,
but it cannot be controlled.
2. Creating a bond in the wrong order can cause the system to fail booting.
- If you create the bond as described above, it inherits the configuration data from the
higher-numbered NIC at creation time. However, upon boot it will try to bond the lower-
numbered NIC first, and that could contain bogus configuration parameters (e.g. zeros)
which will cause unpredictable behavior (potentially a kernel panic/crash).
- The correct procedure for creating a bond is to configure the lowest-numbered NIC (e.g.
eth0) that will be used in that bond, and enslave it FIRST; then enslave any additional
NICs as desired, in numeric order. This will match the behavior at creation time
•There is a known issue when using bonded ethernet interfaces in a clustered environment. The
issue only comes up when using (setNetworkCardChannelBondEnabled T) channel
bonding and all the slave interfaces fail. Normally in an environment where a failover bond is
used, when all the slave interfaces fail, all packages that are using the bond failover to the other
node. However, when the environment uses a channel bond and all the slave interfaces fail, the
packages that are using the bond do not failover to the other node.
•On an HA system, if CIFS is configured to use "shared mode" and a password is applied to a
share, the password will not be mirrored automatically on the other node in the cluster. To
ensure future access to the share when a failover occurs, set passwords ahead of time as
follows:
1. Start the failover package on the first cluster node.
2. Set the share password on that node.
3. Fail the package over to the second cluster node.
4. Set the share password on that node.
5. Fail the package back to the first cluster node.
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 6

•Share passwords must be created in identical order on each node. Share-level security is not
suggested for an HA cluster.
•In an HA environment, the mappings that are done on one node must be duplicated exactly on
the second node. The mappings from one node are not propagated to the second node
automatically. If you fail to have identical mappings on both nodes, users can be prevented from
accessing or modifying files in the event of a failure. For example, assume that a Windows user
that is mapped to a Unix user on Node A has access to a file. If the same mapping does not exist
on Node B and Node A fails, the user could be denied access and other rights to the file. An
alternative to using the mapping function in the user interface is to perform mappings using the
following text API command:
mapClusterSystemWindowsUser(Group)NameToUnixUser(Group)Name
•Snapshots can fail if two or more snapshots are created simultaneously on different nodes of a
cluster. This happens when you schedule the creation of a snapshot for the same time on both
nodes of the cluster. To avoid this problem, stagger the creation time of one of the snapshots by
5 or more minutes.
•If you use the Configuration Wizard to setup your cluster, note that, on the High Availability
Configuration page, the area called Nodes in Cluster at the bottom is not visible until you expand
the dialog vertically.
•The Alerts/Logging sections of Command View NAS 8000 do not propagate the settings from one
node to another. Therefore, you must manage each node individually. This applies to SNMP
Alerts, SMTP E-Mail Alerts, and Remote System Logs.
Client Network Configuration Browser Issues
•Users of Windows 98 clients will need to refresh their Windows Explorer directory view manually
to see changes, additions, or deletions to files contained on the NAS 8000. Windows 2000 and
Windows NT users will not experience this problem.
•The client command showmount does not return the same mount paths as shown in the user
interface. The command will list much longer path names, but they are valid, mountable paths.
For user convenience, we map the longer paths returned by showmount to shorter, easier-to-
remember paths, in the user interface. Either path will mount correctly.
•When accessing the Command View NAS interface, the browser may display a "NOT BOUND IN
REGISTRY" exception, or may appear hung with a "Service not bound" exception displayed in
the Java console. (To view the Java console, click on the Java icon in the Windows system tray,
or in the task list.) This problem may be due to the use of a proxy when accessing the NAS 8000.
If so, configure your browser to bypass the proxy server for the NAS 8000. Another possible
cause is that the Command View NAS graphical user interface (GUI) services need to be
restarted on the NAS 8000 server. To do this, telnet to the NAS 8000 server and login as the
administrator. At the command prompt, type restartSystemGui. You should then be able to
access the Command View NAS interface.
•With the Java 1.4.0 plug-in and Netscape 4.7x, some dialog boxes in the Command View NAS
interface contain un-editable fields after changing the default settings of the dialog box.
- For use with Netscape 4.7x, use the Java 1.3.1 plug-in.
- When an un-editable field is encountered with the Netscape 4.7x/Java 1.4.0 plug-in
combination, click away from (outside of the vicinity of) and back to the problematic dialog
box to force a refresh of the Focus Manager.
•Command View NAS does not work on a Windows 2000 client that is running Windows
NetMeeting 3.0 with remote desktop sharing active. Command View NAS works fine with
NetMeeting, even if remote desktop sharing is enabled, as long as it is not active.
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 7

LUN Management
•Do not use another storage array management tool, such as Command View SDM, to delete
LUNs associated with NAS 8000 volume groups. Once a volume group is created and
associated with a LUN, it is recognized by the NAS 8000 user interface, and must be managed
from it to ensure that you do not lose data.
•If a user accidentally deletes a LUN that was associated with a volume group, and the user
decides to reformat the dedicated-attached array using a text command, the user must follow
these steps:
1. Delete the remaining LUNs on the storage array.
2. Reboot the NAS 8000 server.
3. Execute the following command:
recoverSystemStorageDedicatedAttachConfig
This command will re-create the LUNs, volume groups, and volumes that were previously
being used, prior to the accidental LUN deletion. Caution: Deleting LUNs irreversibly
erases data; use extreme care to avoid data loss.
•As an alternative method to recover from an accidentally deleted LUN, the user may want to
consider deleting the rest of the LUNs in that volume group - leaving the other volume groups and
LUNs alone. In this case, the user should follow these steps:
1. Delete the remaining LUNs.
2. Reboot the NAS 8000 server.
3. Execute the following command:
recoverSystemStorageRegistryConfig
This will recover all storage, except for the volume groups and volumes that were attached to
the deleted LUNs.
4. Manually recreate the deleted LUNs, volume groups, and volumes.
This method preserves most of the user's data - as opposed to wiping out all LUNs by using:
recoverSystemStorageDedicatedAttachConfig.
Volume Group Path Management
•The Test Selected Path menu item on the Volume Group Action menu may not be visible if
you’ve just enabled Data Path Manager in the Applications Tab. If this item is not visible, restart
the browser session and then it will be visible.
•If you choose Test Selected Path and it fails with an error message that says “Handle Existence
Exception”, you can fix the problem by navigating to the Storage tab, Storage Array Summary
page and selecting Actions > Scan for New Storage. Once the scan is completed, Test
Selected Path will work properly.
NFS Performance
•The Linux NFS client implementation in the 2.4.x kernels has a flaw when writing files of size
1024 KB or greater. The NFS caching code in the Linux NFS client is hindered, and the
performance drops to the level of a round trip to the storage subsystem. Testing performed with
HP-UX clients against Linux NFS servers running NFS v3 does not show this flaw.
Insight Manager
•Information about the Emulex Host Bus Adapters (HBAs) is not accessible from the Insight
Manager. However, detailed information about the Emulex HBAs can be retrieved by executing
the following text command:
invokeSystemTool lpfctool –hba –info
•The agents for Insight Manager do provide information about the state of the NAS hardware
platform, including the system board, main memory, I/O boards, and peripheral devices. The
Insight Manager can be accessed from the Status tab, or through a web browser at:
http://<nas machine name_or_ip_address>:2301
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 8

Storage Quotas For Users and Groups
•In quotas, the user/group list is not filtered. Therefore, it is possible to add a user/group quota for
a user/group that already has a quota. In this case, the previous user/group quota is overwritten.
•There is a single grace period per file volume. A grace period set for a file volume will be enforced
on BOTH users and groups regardless of which type of quota was set.
•It is possible that an incorrect user quota size will be reported even when quotas are not enabled.
If a quota is set for a particular user, the size of the soft quota will be reported to CIFS clients as
the size of the disk. The user will be able to write past this reported size as long as there is
actually enough disk space available. To correct this, set the users quota to no quota set. If
using the command line, execute the following command:
setSystemUserQuota <domain+username> <volume> 0 0
Repeat this for any groups that could be demonstrating this behavior.
•When trying to get or set quotas on a system joined to a Windows domain, an error can be
encountered if no users have yet been mapped on the system. This can be fixed by creating a
share and connecting to it, or by explicitly mapping a domain user to a specific UNIX user. As
soon as this occurs, getting and setting quotas are possible.
Snapshot Functionality
•If the NAS 8000 is configured for one time zone and is managed by a client configured for
another time zone, a scheduled event will take place at the client time. For example, if the NAS
8000 is configured for EST and a snapshot is scheduled for 6:00am from a client system
configured for MST, the actual snapshot will occur at 6:00am MST or 8:00am EST. A workaround
is to set the client management system for the same time zone as the NAS 8000. An alternate
workaround is to adjust for the time zone difference and, in this example, set the snapshot for
4:00am from a MST client when you really want a 6:00am EST snapshot.
•The NAS Administrator should carefully plan snapshots (scheduled or manual) to occur
immediately before the creation of the Disaster Recovery files. If a snapshot of file volumes is
taken after creating a DRF and an outage occurs, the snapshot will not be recovered, and any
storage space used by the snapshot will be lost.
•When using the "Snapshot Scheduler" in the Storage tab of HP Command View NAS, the NAS
8000 in a clustered configuration has the following limitation. If the package that owns the volume
for which the snapshot has been scheduled, is not running on the node that originally scheduled
the snapshot, then the snapshot creation will fail. There could be several reasons why the
package and snapshot schedule are not on the same node, such as:
- The package has automatically failed over for some reason, such as NFS service
interruption, or networking hardware failures, and not yet been failed back.
- A deliberate relocation of the package by an administrator who forgot about the snapshot
schedule. There is no built-in way to guarantee that the package owning the volume and the
scheduled snapshot are running on the same node.
•To avoid this, use the command:
setStorageSnapshotStopIoEnabled
•This will temporarily limit access to the NAS 8000 while the snapshot is taken.
•Snapshots taken while the system is under load may cause problems. In environments that are
using NAS OS version 1.6.3-68 in a cluster, using snapshots, and having heavy nfs activity, to
avoid possible corrupt creation of snapshots and subsequent package failures, run the following
command from the command line on both nodes in the cluster.
# setStorageSnapshotStopIoEnabled T
•This will temporarily limit access to the NAS 8000 while the snapshot is taken.
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 9

Adding Storage
•To get a current, accurate representation of the available storage after adding new storage
(arrays or LUNs), click on Actions > Scan for New Storage in Command View NAS. Clicking
the Refresh button on the array summary page will not result in newly added storage being
detected and displayed by the interface.
Deleting Directories
•Directories must be empty prior to being deleted through the Command View NAS interface. If
you attempt to delete a directory that contains files or sub-directories, an error message will
indicate that the command ‘rmdir’ failed. Manually delete any files or sub-directories prior to
deleting the directory though the Command View NAS interface.
Miscellaneous Device Management
•After enabling or disabling an application (for example, Data Path Manager), a Command View
NAS browser session needs to be closed and restarted for the effects of the new application to be
visible in Command View NAS.
•On some occasions, the Exception Handling page under the Support tab will not display any
information in the ‘Debug Level’ border and the Actions menu is disabled. Closing the browser
and restarting Command View NAS will typically resolve this problem.
•In a SAN configuration, the array identifiers are missing on the Command View NAS Identity
page.
•Minimal System requirements for management clients using Sun, Windows and Linux are:
- CPU: Pentium II, 233 MHz or better
- Memory: 128 MB recommended
- Video Card: at least 256 colors
- Note that Command View NAS is a rich client application, so increased memory and CPU
speed along with good network connections will significantly improve the performance of
Command View NAS.
•The Command View NAS interface does not always re-size with the browser window, especially
when running in Mozilla. Restart the browser and re-size it to the desired size before accessing
Command View NAS.
•The Command View NAS interface caches the system time when the user interface is launched.
Subsequent uses of this time from within the user interface will, therefore, be off by the amount of
time that the interface has been up. For example, if the Command View NAS interface is
launched at 1:15pm, and at 1:30pm the administrator creates a snapshot scheduled to expire
after one hour, it will actually expire after 45 minutes. This is because the user interface had a
cached time of 1:15pm when the snapshot was created and scheduled the expiration for 2:15pm
(one hour later) which is really 45 minutes after the snapshot was actually created. There are two
workarounds for this situation:
1. Navigate to Configuration->System Properties, and refresh the page. This will update the
user interface to the current system time.
2. Use the absolute expiration time methodology instead of the relative expiration time
methodology. By setting an absolute time, the snapshot will expire at the desired time as
opposed to a relative offset from the cached system time.
•The “Add Group Mapping Entries” dialog uses the term “User” rather than “Group” for fields in the
dialog. To know if you’re adding group mappings or user mappings, look at the title bar of the
dialog, which is accurate.
Applications
Network Management
•If you are unable to launch Command View NAS from CA Unicenter, you can access Command
View NAS by manually launching your browser and addressing the NAS 8000 directly.
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 10

•The CA Unicenter topology map may report a NAS 8000 as a NAS 8300. Double click on the
icon, select object details, and change the “name” field to NAS 8000.
•In an HA environment, CA Unicenter reports the first package IP address discovered as the IP
address for the cluster.
•The NAS icon does not appear correctly in HP TopTools.
Virus Guard
•The Virus Guard browser interface supports only Netscape and Internet Explorer.
•A volume being scanned cannot be deleted while the scan task is running. Wait until scan task is
complete to delete volume.
•RTP ignores volumes with NFS mount points defined. Only volumes with CIFS shares will be
scanned. Volumes with both NFS mounts and CIFS shares will not be scanned in real time. Use
“Scheduled Scans” to check volumes with NFS mounts points.
•Virus Guard requires volumes to be named with alphanumeric characters only. Volumes named
with special characters will neither display properly nor be scanned. Use the Command View
NAS interface to change the volume names if needed.
•If you use the “Program Files only” option for scan control or real time protection, Virus Guard will
scan a file if it has the execute bit (x-bit) turned on for UNIX systems or if it has an extension that
is known to be susceptible to viruses. The default UNIX attributes for a file written with the CIFS
(Windows) protocol is 766 (rwxrw-rw-). Since the x-bit is set for files whenever CIFS is used to
copy the file to the NAS 8000 server, the “Program Files only” option will end up scanning every
file on the NAS 8000 server. We recommend using the “All Files” option. The "All Files" option
works faster because it bypasses searching the file attributes. Regardless of the option used,
"Program Files only" or "All Files", the result is the same.
•Quarantined files remain in the quarantine section of user volumes even after Virus Guard is
uninstalled. When reinstalled, these quarantined files will show up in the "Manage Quarantined
Files" section of the user interface, as expected. If Virus Guard is to be uninstalled permanently,
all files should be removed from quarantine before Virus Guard is uninstalled. This will release
the space used by quarantined files.
Backup
•The 20/700 tape changer should not be controlled via the NAS 8000, as this is an untested
configuration.
•The NAS 8000 uses a Windows 2000 file permission model, with the default set to “Everyone -
Full Control” with inheritance from the parent. From a NT workstation or server, the default
inheritance cannot be changed unless you install the Microsoft Security Configuration Manager.
This component is available on Service Pack 5 or 6, at (VALUEADD)/mssce/i386/mssce.exe, or
(VALUEADD)/mssce/alpha/mssce.exe. This component provides you with the same permission
control available on the Windows 2000 system.
•The NAS 8000 release 1.6.3 contains the new Data Protector 5.1 agent. An upgrade from NAS
8000 1.6.2 to 1.6.3 does not preserve the agent’s configuration information. The NAS Data
Protector agent should be imported using a Data Protector 5.1 cell master. The NAS-attached
tape devices must be re-configured.
•If NetBackup using the NAS 8000 NDMP service reports a "130 - system call" error when
encountering large inode values, the backup will fail. With NetBackup 4.5 this can be avoided by
setting the NDMP environment variable "SET HIST=P" in the “policy” on sub-tab “file” as the first
line. NetBackup 3.4.1 users can add "SET HIST=N", but only complete session restores are then
possible. We recommend that NetBackup 3.4.1 users upgrade to NetBackup 4.5. Check the
Veritas website for new patches.
•When performing an NDMP backup using the snapshot feature, incremental backups are not
possible. This is because the snapshot file system is read-only, and the file’s “archived” state
cannot be updated.
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 11

•The Windows security data (ACL) is often cached for a SMB connection. If the files are
overwritten by a restore operation modifying the ACL data, the new ACL data may not be visible
until a new SMB connection is established.
Backup Using the OmniBack II 4.1 Agent for NAS 8000
•The NAS 8000 release 1.6.3 contains the new Data Protector 5.1 agent. However, for backup
compatibility the OmniBack II 4.1 agent for the NAS 8000 is available on the NAS support
website. To install this version:
1. Copy the agent to either a CD or to a NAS share. Note that only one agent can be enabled at a time.
Log in as the administrator, and enter the following textApi commands:
osetSystemPackageEnabled hpNas-backupDP F
ouninstallSystemPackage hpNas-backupDP-authorize
ouninstallSystemPackage hpNas-backupOB
2. Install the OmniBack 4.1 agent from either a file or a CD. Enter one of the following sets of commands:
o> installSystemPackage file://vol/x/shared/hpNas-backupOB-4.1-1.i386.rpm
o> installSystemPackage file://vol/x/shared/hpNas-backupOB-authorize.4.1-1.i386.rpm
or
o> installSystemPackage cdrom://hpNas-backupOB-4.1-1.i386.rpm
o> installSystemPackage cdrom://hpNas-backupOB-authorize-4.1-1.i386.rpm
3. To enable the software agent, enter:
o> setSystemPackageEnabled hpNas-backupOB T
Note that installing a new agent does not preserve the agent’s configuration information. The OmniBack II
agent should be imported using the OmniBack II 4.1 or Data Protector 5.0 cell master. The NAS-attached
tape devices must be re-configured.
•The OmniBack II 4.1 agent for the NAS 8000 cannot be used to control older models of HP Tape
Library changers when connected via fibre channel (FC). Connecting tape drives with FC is
supported for models 2/20, 4/40, 6/60, 10/100, 10/180, and 20/700. The Tape Library changer
must be controlled from an NT or HP-UX node (also running Data Protector), when FC-connected
libraries are used. Data Protector with SCSI-connected 2/20, 4/40, 6/60, 10/100, and 10/180
libraries is the preferred configuration.
•When configuring MSL/ESL with Data Protector, the "add device" configuration must be used.
The "auto configure" mode will not correctly associate the tape changer with its tape devices.
Also, the SuperDLT device type is not recognized by the NAS 8000 automatic device configure
module. Both FC and SCSI connections are supported.
•The HP 20/700 tape changer should not be controlled via the NAS 8000, as this is an untested
configuration.
Lotus Domino
•When the NAS 8000 is used as a storage repository for Lotus Domino data, any loss of
connection to the NAS 8000 will require a restart of the server when the storage becomes
available again. To restart the server, on the Lotus Domino server console, type restart server.
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 12

NAS Data Path Manager
LUN Ownership
•If a path failure or series of path failures causes a LUN ownership change between controllers on
an EVA array, failed paths that were originally configured as load balancing paths may be
activated as failover paths when their connectivity is restored. This is because LUNs cannot be
accessed through both controllers at the same time.
VA Array Management Paths
•HP StorageWorks NAS Data Path Manager does not work with paths used to manage HP VA
arrays. These types of paths are only used when the NAS 8000 is in dedicated-attach mode, and
are used to perform array administration tasks such as creating LUNs and reporting on array
health. Depending upon whether the Command View NAS GUI or the command line interface is
being used, you will see one of the following errors:
- In command view NAS:
Error message:
loc: v1.0/Jvm.cpp:xxxx
An error occurred while retrieving data.
Please retry the operation.
.
.
.
Solution:
From Command View NAS at the Storage tab, select Actions > Scan for New Storage.
- From the Command Line Interface
•Error message:
com.hp.HpPanDev.SanDevice.DeviceException: Non Responding Device: c9b0t0l0 -
Unknown SCSI status
at com.hp.HpPanDev.Array2001Obj.CBridge.scsiErrorCheck(
CBridge.java: 8674)
.
.
.
•Solution:
From the Command Line Interface, execute the following command:
% scanStorageHbaScsi
These corrective actions will only work if an alternate management path is available for use
(usually through another HBA). The Data Path Manager product will still automatically switch to
alternate paths to serve data even if a management path is down.
LUN Path Maximum
•Important: There is a 128 logical path limit on the NAS 8000. If you exceed this limit the server
will only detect the first 128 logical paths and disregard any others. Since it is impossible to
predict which 128 logical paths will be seen, this can change between reboots of the system.
HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 13

HP StorageWorks NAS 8000
Version 1.6.3 Release Notes 14
Exceeding the 128 logical path limit may cause unpredictable and undesired effects with your
storage. SAN managers should observe care in how LUNs are presented to the NAS 8000 by
using appropriate zoning techniques.
•A logical path is a physical fibre-channel path from an HBA on the NAS 8000 through zero or
more switches to a port on a controller on a storage array that is presenting the LUN to the HBA.
There are typically 2 logical paths to each LUN on a dedicated attach system that does not use
switches, and 4 logical paths to each LUN on SAN attached systems or dedicated attach systems
using switches.
•Use the following formula to calculate the total number of logical paths: For each LUN: Paths =
(Number of HBA's to which this LUN is presented) * (Number of ports on which this LUN is
presented).
•You can reduce the number of paths by removing LUNs, zoning out ports, or removing cables to
unneeded ports.
•The VA arrays always present LUN 0 for management purposes, even if this LUN has not been
created.
Losing All Paths to a LUN Causes File Volumes to Shutdown
•If all paths to a LUN on a volume group fail, all file volumes on that volume group will shutdown.
Any access to the volumes will result in an I/O error until the volumes are repaired. To repair the
volumes, first restore the paths, then run the repairStorageVolumegroup command on the
affected volume groups. If you still get errors after running the command, reboot the node.
Data Path Manager Installation Process
•There is a known issue that causes the Data Path Manager package to operate without its full
functionality after its initial installation. Specifically, automatic path failback will not be available
immediately after installing and enabling the Data Path Manager software package. This means
that paths that fail and are later restored will not automatically be re-used by the NAS 8000. The
solution to this problem is simply to reboot the NAS 8000 system after the Data Path Manager
package is installed and enabled. Data Path Manager will then operate with its full complement
of functionality, including automatic path failback.
Copyright Hewlett-Packard Co. 2003
This information is subject to change without notice and
is provided "as is" with no warranty.
Hewlett-Packard shall not be liable for any direct,
indirect, special, incidental or consequential damages
in connection with the use of this material