Updated: 11 December 1998

OpenVMS Version 7.2
New Features Manual

Version 3.0 of OpenVMS Management Station adds storage management support to the existing user account and printer management features.

OpenVMS Management Station now makes it easy for you to manage your disk storage devices across multiple OpenVMS Cluster systems and OpenVMS nodes. With Version 3.0 of OpenVMS Management Station, you no longer need to maintain complicated command procedures to control your storage environment. You can create, delete, and manage storage from an easy-to-use Windows interface.

OpenVMS Management Station provides a persistent database that can automatically determine and configure your system's storage configuration at system startup.

OpenVMS Management Station is now based on the Microsoft Management Console (MMC). The Microsoft Management Console provides a common management framework for various administration programs. OpenVMS Management Station is implemented as an MMC snap-in and includes all of the MMC components you need.

3.17.1 OpenVMS Management Station New Features

OpenVMS Management Station Version 3.0 includes the following new features:

• Storage Configuration Management. You can create, delete, and manage a disk volume under one interface. Wizards make it easy to create a volume.
• Configuration Control Over Reboots. If you allow it to, OpenVMS Management Station preserves the disk configuration across reboots. OpenVMS Management Station can mount and maintain your storage configuration without intervention.
  And, OpenVMS Management Station mounts volumes much faster than is possible with DCL or command procedures, so performance is enhanced.
• Coexistence with Established Environment. You do not have to change your existing DCL command procedures if you do not want to. You can use the display capabilities of OpenVMS Management Station without having to use the automated mount feature. In this way, you can give OpenVMS Management Station as little or as much control as you are comfortable with.
• Remote Management Supported. You can use OpenVMS Management Station to remotely manage your OpenVMS systems. Once you establish a TCP/IP dialup connection, you can then use OpenVMS Management Station to manage your OpenVMS systems from home, while traveling, and so forth.
• Integrated Web Features. MMC allows you to insert World Wide Web URLs directly into the viewer and access them via Microsoft Internet Explorer. OpenVMS Management Station includes hot links to the OpenVMS and OpenVMS Management Station web pages, where you can find the latest product information and kits. You can use this feature to add web links of your own. For example, you might add a link to an internal Help Desk troubleshooting page, or to a system management procedures page.

To allow OpenVMS and Windows NT to interoperate, Compaq has provided a registry on OpenVMS. Both PATHWORKS and COM for OpenVMS use the OpenVMS Registry. Like the Windows NT registry, the OpenVMS Registry is made up of two components: the OpenVMS Registry database and the OpenVMS Registry server. The OpenVMS Registry database is a systemwide or clusterwide hierarchical database of configuration information. This information is stored in a database structure of keys and associated values. The OpenVMS Registry server controls all OpenVMS Registry operations, such as creating and backing up the OpenVMS Registry database, and creating, displaying, modifying, or deleting keys and values.

The OpenVMS Registry includes interfaces (COM APIs and system services) to allow applications to control the OpenVMS Registry server and to read and write to the OpenVMS Registry database. The OpenVMS Registry also includes server management utilities to allow system managers to display and update OpenVMS Registry information from the OpenVMS DCL command line.

The OpenVMS Registry is compatible with the Windows NT registry. Windows NT client applications such as RegEdt32 can connect to and edit the OpenVMS Registry.

3.18.1 For More Information

For more information about the OpenVMS Registry, see the OpenVMS Connectivity Developer Guide. You can find this document in the following locations:

• The document is included in the COM for OpenVMS kit. When you install the kit, the document is placed in the COM for OpenVMS [SYSHLP] subdirectory.
• The document is available from the OpenVMS website at the following location:

  http://www.openvms.digital.com/openvms/products/dcom/index.html

For more information about installing and managing the OpenVMS Registry, refer to the OpenVMS System Manager's Manual.

3.19 PCI Gigabit Ethernet Support (Alpha Only)

Run-time support for the DIGITAL PCI-to-Gigabit Ethernet adapter (DEGPA) is included in OpenVMS Alpha Version 7.2. The DEGPA is supported on single systems. For information about OpenVMS Cluster support, see Section 3.16.3.

The DEGPA conforms to the IEEE 802.3z Gigabit Ethernet standard running over 1000BASE-SX fiber optic cabling. It can run point-to-point to another Gigabit Ethernet adapter or to a Gigabit Ethernet hub or switch.

To configure the DEGPA, do the following:

1. Verify that your system is running OpenVMS Alpha Version 7.2.
2. Power down your system.
3. Install the DEGPA in an available PCI slot, preferably a 64-bit slot. (For installation instructions for this adapter, see the DIGITAL PCI-to-Gigabit Ethernet Adapter Installation and Configuration guide that accompanies it.)
4. Boot the system.
5. Determine which EW device is the DEGPA with the following command:

   $ RUN SYS$SYSTEM:LANCP LANCP> SHOW DEV EW/CHAR

   
   The device that has a line speed of 1000 megabits per second is the DEGPA.

6. Verify that the device is cabled correctly to the switch or point-to-point with another device. If it is, a "link up" console message is displayed. If you missed the console message (it is not logged to the OPCOM log file), you can either check the DEGPA or use the SDA utility:
   • On the DEGPA, there are two small lights. The light next to the link symbol (two circles linked together) is green if the DEGPA is correctly cabled. (For more information about the lights, see the DIGITAL PCI-to-Gigabit Ethernet Adapter Installation and Configuration guide.)
   • Issue the SDA command SHOW LAN, as follows:

     $ ANALYZE/SYSTEM SDA> SHOW LAN

     
     If cabled properly, SDA displays "ring available" for the device.

As firmware is updated for the Gigabit Ethernet adapter, it will be included in successive versions of the driver. Unlike other adapters, the firmware image for the Gigabit Ethernet adapter is embedded in the driver. (The firmware is loaded into the adapter at device initialization and is part of the driver image.) This means that you do not update the firmware.

3.19.2 Link Autonegotiation Support

The OpenVMS Gigabit Ethernet driver supports link autonegotiation. Autonegotiation is enabled by default and can be turned on or off using the SYSMAN utility or the LANCP utility. If you use the SYSMAN utility to change the setting, it takes effect on all the Gigabit Ethernet adapters. The LANCP utility can be used to change the setting on individual adapters or on all of them. The change takes effect immediately when issued by either utility.

Many Gigabit Ethernet switches support link autonegotiation but some, such as the Gigaswitch/Ethernet, do not. If the Gigabit Ethernet switch you are using does not support link autonegotiation, you must disable link autonegotiation on the driver.

Each time the setting of link negotiation is changed, a console message is displayed. At driver initialization, if you have disabled autonegotiation, the following message is displayed on the console:

%EWA0, Autonegotiation disabled per SYSGEN parameter LAN_FLAGS

Because the Gigabit Ethernet driver checks the LAN_FLAGS setting once every second and changes the link setting as dictated by bit 5, the driver will be able to see a valid link within a few seconds.

If the device is cabled properly but the driver is not able to see a valid link, it is probably because the switch does not support autonegotiation.

Directions for using SYSMAN or LANCP to turn autonegotiation off or on follow.

3.19.2.1 SYSMAN Utility Directions

Using SYSMAN, you can disable or enable autonegotiation on all adapters. You cannot single out individual adapters. To disable autonegotiation, you set bit 5 of the LAN_FLAGS system parameter. To enable autonegotiation, you clear the bit.

To disable autonegotiation:

1. Verify that the device is cabled properly by looking at the link light or by using SDA, as previously described.
2. Invoke SYSMAN and issue the following SYSMAN commands:

   $ run sys$system:sysman SYSMAN> parameters use current SYSMAN> parameters set lan_flags %x20 SYSMAN> parameters write active SYSMAN> parameters write current SYSMAN> exit

   
   The notation "%x" instructs the operating system that the following number is hexadecimal.

   Note The setting for bit 5 should be OR'd with any other existing settings for the LAN_FLAGS system parameter.

3. Insert or edit the following line in SYS$SYSTEM:MODPARAMS.DAT:

   lan_flags = %x20 ! turn off auto-negotiate on DEGPA

To enable autonegotiation, follow the previous directions except clear bit 5 instead of setting it. Then modify the relevant line in SYS$SYSTEM:MODPARAMS.DAT.

3.19.2.2 LANCP Utility Directions

You can use the LANCP utility to disable or enable autonegotiation on one or more specific adapters with separate commands or on all the adapters with a single command.

Issue the following command to disable link autonegotiation for a specific adapter, for example, ewa:

$ mc lancp set dev devname/noauto ewa

Issue the following command to disable link autonegotiation on all Gigabit Ethernet adapters connected to the Gigabit Ethernet switch:

$ mc lancp set dev devname/noauto ew/all

If you have disabled link autonegotiation, you can turn it back on by using the syntax of the previous commands but replacing noauto with auto.

3.19.3 Jumbo Frame Support

The OpenVMS Gigabit Ethernet driver supports frame sizes larger than 1518 bytes, the maximum frame size supported on Ethernet and Fast Ethernet. The Gigabit Ethernet adapter supports frame sizes up to 9018 bytes.

The Gigabit Ethernet driver limits the actual maximum frame size to the size of the non-paged pool lookaside lists less the transmit and receive request frame overhead of 640 bytes. Table 3-11 shows the default frame size and the maximum size allowed for this release and for OpenVMS Version 7.2.

Table 3-11 OpenVMS Frame Size Support

Version	Default Size	Maximum Size
OpenVMS Version 7.1--2	1512 bytes	4480 bytes
OpenVMS Version 7--2	1512 bytes	7552 bytes

Some Gigabit Ethernet switches, such as the Gigaswitch/Ethernet, do not support frame sizes larger than 1518 bytes. Other switches may support the larger frame size, but may require applications to determine whether the larger frame size can be used when communicating with another node on the same switch or elsewhere on the LAN. For example, the switch may accept large frames but may discard them, without notice to the sender, when the destination is a Fast Ethernet node on the same switch.

Each application is responsible for negotiating a frame size between nodes. Some applications, such as Digital TCP/IP, rely on the switch for frame fragmentation or notification of frame size errors (ICMP frame for TCP/IP).

If the switch does not provide jumbo frame support, the maximum frame size must be limited by the LAN driver to 1518 bytes. This limit enables communication over the switch to nodes that do not support the larger frame size or that are limited by switch hardware between two nodes.

You can use the SYSMAN utility or the LANCP utility to enable the use of jumbo frames. Directions for using both utilities follow.

The jumbo frame setting takes effect immediately, regardless of which utility you use. However, an application may check the maximum packet size only during initialization. For this reason, it may be necessary to stop and restart the application to effect the change.

Each time you change the frame size, a console message is displayed. If you change the frame size with SYSMAN, the following message is displayed:

%EWA0, Jumbo frames enabled per SYSGEN parameter LAN_FLAGS

3.19.3.2 SYSMAN Utility Directions

To enable jumbo frames by means of the SYSMAN utility:

1. Set bit 6 of the LAN_FLAGS system parameter, as shown in the following example:

   $ run sys$system:sysman SYSMAN> parameters use current SYSMAN> parameters set lan_flags %x40 SYSMAN> parameters write active SYSMAN> parameters write current SYSMAN> exit

   
   The notation "%x" instructs the operating system that the following number is hexadecimal.

   Note The setting for bit 6 should be OR'd with any existing bits set in LAN_FLAGS.

2. Insert or edit the following line in SYS$SYSTEM:MODPARAMS.DAT:

   lan_flags = %x40 ! enable jumbo frames on DEGPA

You can turn off the jumbo frame setting by clearing bit 6 and then editing the same line in SYS$SYSTEM:MODPARAMS.DAT.

3.19.3.3 LANCP Utility Directions

To enable jumbo frames by means of the LANCP utility:

1. Issue the following command:

   $ mc lancp set dev devname/jumbo_frames

2. Insert the previous command into SYS$SYSTEM:SYSTARTUP_VMS.COM to enable jumbo frames before applications are started.

To turn off jumbo frame support, replace jumbo_frames with nojumbo_frames and make the appropriate edit to the SYS$SYSTEM:SYSTARTUP_VMS.COM file.

3.20 PCSI Qualifiers

Beginning with OpenVMS Version 7.2, the POLYCENTER Software Installation (PCSI) utility has two new qualifiers that you can use with a number of PRODUCT commands:

Qualifier	Description
/KIT_ATTRIBUTES	Allows you to select kits by kit type or kit format, or both.
/SPAN_VERSIONS	Allows you to specify a range of versions using the keywords ABOVE, BELOW, MAXIMUM, and MINIMUM.

3.21 RMS CONVERT Utility Enhancements

For OpenVMS Version 7.2, the CONVERT utility has had several enhancements added to it. These improved features are described in the following paragraphs. For more detailed information, see the OpenVMS Record Management Utilities Reference Manual and the OpenVMS Utility Routines Manual.

The CONVERT utility has been modified to eliminate a previous design constraint in which the output file would temporarily become vulnerable to user access during the exchange of the file between CONVERT and the SORT32 utility. This would occur during the FAST load processing of the secondary keys of the file.

Previous versions of CONVERT required unlocking of the output file whenever it passed control to the SORT32 utility for performing sort operations. This provided an opportunity in which the output file could be locked by a user application, causing the file to become inaccessible to CONVERT. This could result in a premature exit of the convert process, leaving some of the secondary keys in an uninitialized state. This failure was most prevalent when the output file name was the same as the production file name.

The CONVERT modifications were made with the goal of not impacting performance or increasing disk space usage by default. In all test cases, the CONVERT performance has been either equal to or better than previous versions of CONVERT while maintaining constraints on the disk space required for the work files. A new qualifier, /SECONDARY, has been added that allows a user to override the latter default constraint to improve performance at the expense of disk space.

3.22 SYSMAN RESERVED_MEMORY: New and Modified Subcommands and Qualifiers

If you want to reserve a block of contiguous page frame numbers (PFNs), you can use new and modified SYSMAN RESERVED_MEMORY subcommands to do so. The following sections discuss:

• Changes to existing subcommands and qualifiers
• New subcommands and qualifiers that have been added

The following sections lists new and modified qualifiers of existing RESERVED_MEMORY subcommands.

New Qualifiers /SYSGBL and /[NO]GLOBAL

The subcommands ADD, FREE, REMOVE, and SHOW accept these new qualifiers:

/SYSGBL	Specify this qualifier to indicate that the reservation is for a system global memory resident section.
/GLOBAL_SECTION /NOGLOBAL_SECTION	Specify /NOGLOBAL_SECTION to indicate that the memory qualifier is for a privileged application rather than for a group or system global section.

Changes Related to the /GROUP Qualifier

You cannot use the /GROUP qualifier with either /SYSGBL or /NOGLOBAL_SECTION.

Improved Ease of Use of /ZERO and /ALLOCATE Qualifiers

In OpenVMS Version 7.1, you could specify /ZERO only if you also specified /ALLOCATE. The new behavior is the following:

• /ZERO implies /ALLOCATE.
• /NOALLOCATE implies /NOZERO.
• /ZERO is incompatible with /NOALLOCATE.

This change has no effect on existing command procedures because all commands that were valid in Version 7.1 behave exactly as they did in Version 7.1 because /GLOBAL_SECTION is now the default.

3.22.2 New MODIFY Subcommand and Qualifiers

OpenVMS Alpha Version 7.2 introduces several new system parameters. Table 3-12 lists the parameters and their functions. For additional information on OpenVMS system parameters, refer to online help and the OpenVMS System Management Utilities Reference Manual.

Table 3-12 New System Parameters

Parameter	Function
ARB_SUPPORT	A number of COMPAQ and third-party products require that OpenVMS maintain the pre-Version 7.2 Access Rights Block (ARB) and related kernel-mode data structures in parallel with the data structures that have replaced the ARB; this requirement continues because many products have not yet been modified to operate correctly with the new per-thread security Persona Security Block (PSB) data structures.
MAXBOBS0S1	On Alpha systems, MAXBOBS0S1 defines the maximum amount of 32-bit system space, measured in pagelets, that can be used for windows to buffer objects. MAXBOBS0S1 is a DYNAMIC parameter.
MAXBOBS2	On Alpha systems, MAXBOBS2 defines the maximum amount of 64-bit system space, measured in pagelets, that can be used for windows to buffer objects. MAXBOBS2 is a DYNAMIC parameter.
MPDEV_D1	MPDEV_D1 is reserved for Compaq use only.
MPDEV_ENABLE	MPDEV_ENABLE controls the formation of multipath sets. When MPDEV_ENABLE is ON (1), the formation of multipath sets is enabled. Setting this parameter to OFF (0) disables the formation of additional multipath sets; path failover, however, continues on existing multipath sets. The default value is ON (1). MPDEV_ENABLE is a DYNAMIC parameter.
MPDEV_LCRETRIES	MPDEV_LCRETRIES specifies the number of times the system retries locally connected paths to a SCSI device before moving on to local unconnected paths or to the MSCP-served path to the device. The valid range is 1 through 256. MPDEV_LCRETRIES is a DYNAMIC parameter.
MPDEV_POLLER	MPDEV_POLLER controls the polling of paths to multipath set members. Setting this parameter to ON (1) allows early detection of errors on otherwise inactive paths. If a path becomes unavailable or returns to service, an OPCOM message notifies the system manager. If MPDEV_POLLER is set to OFF (0), multipath polling is disabled. The default value is ON (1). MPDEV_POLLER is a DYNAMIC parameter.
MPDEV_REMOTE	MPDEV_REMOTE controls whether the MSCP served path to a SCSI device can become a member of a multipath set. Setting the parameter to ON (1) allows this membership. If MPDEV_REMOTE is set to OFF (0), only local paths to a SCSI device can be used to form additional multipath sets; path failover, however, continues to include remote members of existing multipath sets. The default value is ON (1). MPDEV_REMOTE is a DYNAMIC parameter.
POWEROFF	POWEROFF enables or disables software requests to the console firmware to remove power from the system. This parameter should normally be turned ON (1) to allow software to make power-off requests. However, POWEROFF can be set to OFF (0) to disable software power-off requests. If firmware or hardware support for the power-off request is not implemented, the shut-down procedure leaves the system halted, but fully powered. POWEROFF is a DYNAMIC parameter.
NPAG_BAP_MIN_PA	On Alpha systems, NPAG_BAP_MIN_PA specifies the lowest physical address allowed within bus addressable pool.
VCC_MAX_CACHE	On Alpha systems, VCC_MAX_CACHE is reserved for Compaq use only.
VCC_MAX_LOCKS	On Alpha systems, VCC_MAX_LOCKS is reserved for Compaq use only.
VCC_MINSIZE	On VAX systems, VCC_MINSIZE sets the lower limit in pages of memory used by virtual I/O cache.
VCC_WRITEBEHIND	On Alpha systems, VCC_WRITEBEHIND is reserved for Compaq use only.
VCC_WRITE_DELAY	On Alpha systems, VCC_WRITE_DELAY is reserved for Compaq use only.
WLKSYSDSK	On Alpha systems, WLKSYSDSK is used by various bootstrap components to determine if the system disk should be treated as though it is write-locked.

Copyright © Compaq Computer Corporation 1998. All rights reserved. Legal

6520PRO_004.HTML