HP OpenVMS Systems Documentation

HP OpenVMS Version 8.3 New Features and Documentation Overview

This chapter provides information about new features for all users of the HP OpenVMS Alpha and OpenVMS for Integrity servers operating systems.

2.1 New Integrity Server Support

OpenVMS Version 8.3 supports new entry-level, midrange, and high-end Integrity server systems. In addition, OpenVMS Version 8.3 continues to support all of the Integrity servers and options supported in OpenVMS Version 8.2 and Version 8.2--1.

2.2 Batch Queue Job Limit Increased

In OpenVMS Version 8.3, the maximum job limit on batch queues is increased from 255 to 65535.

2.3 DCL Commands and Lexical Functions

Table 2-1 and Table 2-2 summarize new and changed DCL commands, qualifiers, and lexical functions for OpenVMS Version 8.3. A few additional new features related to DCL usage are described in the following sections. For more information, see the online help or the HP OpenVMS DCL Dictionary.

You can define the new symbol DCL$CTRLT_PID to point to a remote process ID. If you have the necessary privileges, you can then display Ctrl/T information for that process. The remote process can be a different process on the same system or a process on a different system in the cluster. The following example shows how a privileged user on NODE1 can define DCL$CTRLT_PID to view process information for user JSMITH on NODE2.

$ <Ctrl/T>
NODE1::SYSTEM 17:40:55   (DCL)   CPU=00:00:00.16 PF=212 IO=98 MEM=146
$
$ DCL$CTRLT_PID="23800436"  !Define symbol to point to remote process ID
$
$ <Ctrl/T>
NODE2::JSMITH 17:41:12 LOOPER    CPU=01:28:05.17 PF=2700 IO=594 MEM=322
$

On an image rundown, DCL populates the $SEVERITY and $STATUS symbols. With Version 8.3, two new symbols are added, $FACILITY and $IDENT, which contain the facility number and the message number.

$ EXIT  %X10911A02
$ SHOW SYMBOL $STATUS
  $STATUS == "%X10911A02"
$ SHOW SYMBOL $FACILITY
  $FACILITY == "%X00000091"
$ SHOW SYMBOL $IDENT
  $IDENT == "%X00000340"
$ SHOW SYMBOL $SEVERITY
  $SEVERITY == "2"

By defining the new symbol DCL$CTRLT, users can augment the traditional Ctrl/T output with the text defined for DCL$CTRLT. This technique can be used to indicate progress in user applications or for debugging purposes. The following example demonstrates the use of DCL$CTRLT within a command procedure. The command procedure is running in a loop that updates the symbol DCL$CTRLT to indicate the number of loop iterations executed so far.

$ TYPE CTRLT_LOOP.COM
$ inner=0
$ outer=0
$ loop:
$ loop1:
$ if inner .gt. 20000 then goto end_loop1
$ inner=inner+1
$ dcl$ctrlt=F$FAO("Inner loop count is !SL !/ -
_$ Outer loop count is !SL",inner,outer)
$ goto loop1
$ end_loop1:
$ inner=0
$ outer=outer+1
$ goto loop
$
$ @CTRLT_LOOP
NODE1::JSMITH 10:46:37   (DCL)   CPU=00:03:42.68 PF=13453 IO=6743 MEM=187
Inner loop count is 12306
 Outer loop count is 0
NODE1::JSMITH 10:46:43   (DCL)   CPU=00:03:49.19 PF=13455 IO=6744 MEM=187
Inner loop count is 19200
 Outer loop count is 2
 .
 .
 .

You can now specify /SINCE=JOB_LOGIN in any command that accepts the /SINCE qualifier (except SHOW LICENSE). JOB_LOGIN refers to the login time of the master process in the job. For example, PIPE creates a subprocess for each pipe segment; therefore /SINCE=LOGIN cannot be effectively used in a pipe.

$ PIPE DIRECTORY/SINCE=LOGIN | SEARCH SYS$INPUT TEST
%SEARCH-I-NOMATCHES, no strings matched
$
$ PIPE DIRECTORY/SINCE=JOB_LOGIN | SEARCH SYS$INPUT TEST
TEST.TXT;1

The maximum per-I/O block count supported for the COPY command has been increased to 2**31 --1. Note that the block count is reduced if necessary to the maximum supported by the device driver performing the I/O. For example, the maximum I/O size supported by SCSI Fibre Channel drivers is currently 256 blocks.

2.3.6 Increase Maximum Prompt Size

The maximum DCL prompt size has been increased from 32 characters to 64 characters to accommodate special prompts and escape sequences, which some users require.

2.4 Hyper-Threading (I64 Only)

OpenVMS for Integrity servers Version 8.3 supports Hyper-Threading with nPartitions on dual-core Intel Itanium 2 processors. Hyper-Threading provides the ability for processors to create a second logical CPU on a single core that may allow additional efficiencies of processing. For example, a dual-core processor with Hyper-Threading active provides four logical CPUs, two on each core.

The EFI Shell cpuconfig command can enable and disable Hyper-Threading for an nPartition whose processors support it. Recent releases of the nPartition Command and Partition manager also support Hyper-Threading.

The effect that hyperthreads have on performance depends heavily on the application mix that is running. HP recommends that you start with hyperthreads turned off and experiment later. Two CPUs that share a core when hyperthreading is enabled are referred to as cothreads. The SHOW CPU/BRIEF and SHOW CPU/FULL commands now provide information about cothreads. For example:

$ show cpu/brief 3

System: XYZZY, HP rx4640

CPU 3    State: RUN                 CPUDB: 820DB480     Handle: 000060A0
         Owner: 000004CB8         Current: 000004C8     Partition 0 (XYZZY)
         COTHd:          1

In this example, COTHd: 1 indicates that CPU 3 and CPU 1 share the same core.

A sample cpuconfig command output looks like this:

Shell> cpuconfig

PROCESSOR MODULE INFORMATION

        # of               L3      L4      Family/
CPU     Logical            Cache   Cache     Model          Processor
Module     CPUs   Speed    Size    Size     (hex.)    Rev   State
-----   -------   ------   ------  ------  ---------  ---   -----
 0           4    1.4 GHz   6 MB   None    20/00      CO    Active

CPU threads are turned on.

For information about the cpuconfig command, see the HP OpenVMS Version 8.3 Upgrade and Installation Manual. For information about Hyper-Threading and nPartitions, see the HP System Partition Guide: Administration for nPartitions.

2.5 HP Instant Capacity (iCAP) and HP Temporary Instant Capacity (TiCAP) (I64 Only)

OpenVMS Version 8.3 now supports iCAP, which is the HP Utility Pricing Solutions product for cell-based Integrity servers that has a pricing model based on purchasing components (processors, cell boards, and memory). With Instant Capacity you initially purchase a specified number of activated components and pay a Component without Usage Rights (CWUR) fee for a specified number of deactivated components. To activate a component, you pay the balance of the component price and license it through the application of a codeword obtained from the secure iCAP web portal. Components can be instantly activated without requiring a reboot.

OpenVMS Version 8.3 also supports TiCAP, an HP product that enables customers to purchase prepaid processor-activation rights for a specified (temporary) period of time. Temporary capacity is in increments, such as 20-day or 30-day increments, where a day equals 24 hours for a core.

For more information about iCAP and TiCAP, see the HP Instant Capacity User's Guide, at the following Web site:

The following sections describe the changes and enhancements to the License Management Facility.

2.6.1 LMF Compliance Report

You can now change the target account for LMF compliance reports. LMF sends the compliance report to the account defined by the logical name LMF$COMPLIANCE_CONTACT_ACCOUNT, but the default account is still the SYSTEM account.

2.6.2 License Terminology Change (I64 Only)

The licensing and terminology for OpenVMS for Integrity servers has changed from Per Processor License (PPL) to Per Core License (PCL). With the introduction of the dual-core Intel Itanium 2 processor, the term "processor" does not mean what it meant previously. With dual-core Intel Itanium 2 systems, each processor contains 2 cores, and these systems are licensed by the number of active cores. On a system other than a dual-core Intel Itanium 2 system, core equals processor.

PCL implements the licensing model on OpenVMS for Integrity servers. In the PCL model, a product is licensed according to the number of active processor cores on the system, not the static rating scheme used for Alpha and VAX systems. Each active processor core requires one PCL unit. If you increase or decrease the number of active processor cores on a system, the requirement for PCL licenses changes.

A PCL license is required to run operating environments, OE products purchased separately (like clustering), and many standalone products on OpenVMS for Integrity servers.

PCL licenses offer flexibility because you can purchase licenses in the exact number you need, and you can move the licenses to other processors. If you upgrade or reconfigure your system with additional processor cores, you purchase additional PCL licenses.

LMF constantly checks the number of PCL licenses against the number of active processor cores and enforces a soft compliance model. Any changes to the system are noted and are checked for compliance.

To implement PCL, the Hardware_ID option CPU_SOCKETS=n is changed to SOCKETS=n. Also, the SHOW LICENSE/CHARGE command is updated to display the number of active cores on the system.

The HP OpenVMS License Management Utility Manual has not been updated for this release with the new terminology. As you read the manual, take note of the following terminology changes:

• Per Processor License is now Per Core License.
• PPL is now PCL.
• CPU is now processor core.
• CPU_SOCKETS=n is now SOCKETS=n.

Note the following definitions:

• Processor --- The component that plugs into a processor socket. The processor can contain more than one processor core.
• Processor module --- The packaging of one or more processors to connect into a single socket on a system bus.
• Core --- The actual data-processing engine within a cell-based processor. A single processor can contain multiple cores.
• Processor socket --- The system board socket into which a processor attaches.

The HP nPartition Provider is now ported to OpenVMS Version 8.3 in support of Instant Capacity (iCAP) features on cell-based Integrity servers.

Note, the OpenVMS Version 8.3 nPartition Provider does not support local or remote nPartition management tasks, such as remote WBEM connections to the nPartition Provider running on OpenVMS Version 8.3.

2.8 HP Pay per Use (PPU) (I64 only)

OpenVMS Version 8.3 supports Pay per use for cell-based Integrity server systems leased through HP Finance. PPU enables customers to pay for only the processing cycles that they actually consume. There are two types of PPU:

• Percent CPU
  Continuously monitors the utilization of every CPU in the system.
• Active CPU
  Continuously counts the number of active CPUs in the system.

The system manager can instantly activate inactive CPUs to cope with additional loads. With both Percent CPU and Active CPU, the utilization data is sent to a separate Utility Meter and then to the secure HP PPU Web portal, where you can obtain various usage reports within 48 hours. (Note that Instant Capacity and Pay per use are mutually exclusive on any cell-based Integrity server.)

For more information about PPU, see the HP Pay per use (PPU) User's Guide, at the following Web site:

HP Superdome servers based on the HP sx1000 chipset can support configurations with both PA-RISC nPartitions and Intel Itanium 2 nPartitions in the same server.

Specific hardware, firmware, operating systems, and management tools are required for including both PA-RISC nPartitions and Intel Itanium 2 nPartitions on Superdome hybrid servers.

The HP OpenVMS for Integrity servers Version 8.3 release is supported on Superdome hybrid servers in an nPartition with the Intel Itanium 2 single-core processor with 9 MB cache.

For details and requirements, see the HP Superdome Hybrid Servers: Intel Itanium 2 and PA-RISC nPartition Mixing on the following Web site:

This document is available in the Systems Hardware area, under the HP Integrity Superdome Server and HP 9000 Superdome Server headings.

2.10 HP Web-Based Enterprise Management Services for OpenVMS (WBEM)

WBEM is an optional product available on HP OpenVMS for Integrity server systems that provides an industry-standard enterprise management framework and resource description. The WBEM structured framework is capable of being extended and uses Internet standards. Developers of management applications can take advantage of work previously done to expose resource information and operations. For example, developers can take code that was developed for a specific platform or application and use it with WBEM for the same purposes.

This chapter provides information about new features, changes, and enhancements for system managers.

3.1 BACKUP Utility Enhancements

OpenVMS Version 8.3 contains several enhancements to the Backup utility:

• Dynamic volume expansion (DVE)
• Save-set encryption
• A more complete Ctrl/T message displayed during an interactive backup operation
• A new /PROGRESS_REPORT qualifier that sends the expanded BACKUP messages to the current output device
• A new /IO_LOAD qualifier that affects the number of simultaneous READ I/Os on your system

For more information about all of these enhancements, see the HP OpenVMS System Management Utilities Reference Manual.

3.1.1 Dynamic Volume Expansion (DVE) Support in OpenVMS Backup Utility

Beginning with OpenVMS Version 8.3, volume expansion size is recorded in the save-set header when you perform image backup to a save set. Previously, BACKUP had no knowledge of DVE when initializing an output volume. This meant that, in performing disk-to-disk backup or restoring a save set, BACKUP did not preserve the volume expansion size or the logical volume size.

3.1.1.1 Volume Expansion Size

BACKUP/LIST now displays the volume expansion size if it exists in the save set. When you restore a save set (or perform a disk-to-disk backup), the target device inherits the volume expansion limit from the save set. If the save set does not include expansion size, BITMAP.SYS is sized as it was in previous versions of OpenVMS.

/IGNORE=LIMIT Option

A new option, /IGNORE=LIMIT, prevents the target device from inheriting the expansion limit.

/LIMIT Qualifier

The new qualifier /LIMIT allows you to specify the expansion size limit during restore or save operations regardless of the value stored in the save-set header, which matches how the /LIMIT qualifier of the INITIALIZE utility works.

3.1.1.2 Logical Volume Size

By default, the logical volume size is not preserved because restoring a save set of 2GB to a 4GB disk, for example, would result in only 2GB of available disk space.

/SIZE Qualifier

To preserve the logical volume size, use the /SIZE qualifier. When you specify /SIZE, the geometry of the target device is determined by the logical size rather than calling $GETDVI to find out the physical limits of the device.

The /SIZE qualifier uses an optional value as the new logical size of the target device. This new value overrides any existing value in the save set, which matches how the /SIZE qualifier in the INITIALIZE utility works.

/NOINITIALIZE Qualifier

A restriction in using BACKUP/NOINITIALIZE is that this command does not preserve the DVE characteristics of the output device. The reason is that the target device is mounted foreign, preventing OpenVMS from obtaining the expansion size and the logical size. To overcome this restriction, use the /LIMIT and /SIZE qualifiers.

The chapter "Managing Storage Media" in HP OpenVMS System Manager's Manual contains a section that explains DVE in more detail.

3.1.2 Encrypting BACKUP Save Sets

The OpenVMS Backup utility provides protection against file or volume corruption by creating functionally equivalent backup copies. BACKUP creates save sets, which are written in BACKUP format so that only BACKUP can interpret the data in a save set. When you create save sets, you can protect them further by encrypting them.

OpenVMS Version 8.3 includes the following new encryption features:

• Support for Advanced Encryption Standard (AES) encryption algorithms
• Support for the following AES keywords for the ALGORITHM option:

  AESCBC AESCFB AESECB AESOFB

  
  Each of these modes can have a user-defined secret key in one of three different lengths (128, 192, and 256) for a total of 12 possible key combinations. If you specify =AES after the ALGORITHM option, the default is AESCBC128.
• Ability to encrypt the data in a save set using the AES encryption algorithm specified by the user.