| Previous | Contents | Index |
The following features are added to the LAN-based cluster communications driver (PEdriver) in in OpenVMS Version 8.3:
Data compression may be used to reduce the time to transfer data between two OpenVMS nodes when the LAN speed between them is limiting the data transfer rate, and there is idle CPU capacity available. For example, it may be used to reduce shadow copy times, or improve MSCP serving performance between Disaster Tolerant Clusters sites connected by relatively low-speed links such as E3 or DS3, FDDI, or 100Mb Ethernet. PEdriver data compression can be enabled by using SCACP, Availability Manager, or the NISCS_PORT_SERV sysgen parameter.
The number of packets in flight between nodes needs to increase proportionally to both the speed of LAN links and the inter-node distance. Historically, PEdriver had fixed transmit and receive windows (buffering capacity) of 31 outstanding packets. Beginning with OpenVMS Version 8.3, PEdriver now automatically selects transmit and receive window sizes (sometimes called pipe quota by other network protocols) based on the speed of the current set of local and remote LAN adapters being used for cluster communications between nodes. Additionally, SCACP and Availability Manager now provide management override of the automatically-selected window sizes.
For more information, see the SCACP utility chapter, and
NISCS_PORT_SERV in the HP OpenVMS System Management Utilities Reference Manual and the HP OpenVMS Availability Manager User's Guide.
3.9 OpenVMS Operating System Media Patch-Related Menu Option
The OpenVMS operating system distribution media main menu now includes a new option (7) that enables you to perform patch-related operations. When you select option 7, you are brought to a submenu that provides options enabling you to search the patch kits, install patches, remove recent patches for which there is recovery data, and to show and delete recovery data. As a result, you can perform these operations even when the operating system cannot be booted (in which case you cannot use the PCSI PRODUCT command). This example shows the menu options:
Please choose one of the following: 1) Upgrade, install or reconfigure OpenVMS I64 Version X8.3-BBV 2) Display layered products that this procedure can install 3) Install or upgrade layered products 4) Show installed products 5) Reconfigure installed products 6) Remove installed products 7) Find, Install, or Undo patches; Show or Delete Recovery Data 8) Execute DCL commands and procedures 9) Shut down this system Enter CHOICE or ? for help: (1/2/3/4/5/6/7/8/9/?) |
Although options 2 and 3 of the main menu can still be used to perform
patch operations, HP recommends that you use the new option 7 submenu
operations; the functionality provided by these operations is more
extensive and more reliable. For example, when you install patches
using the new submenu option, recovery data is saved automatically. In
addition, the new options allow you to specify locations on which to
perform the operations (in addition to the default target device). You
can use wildcards to specify locations.
3.10 PCSI Utility Enhancements
OpenVMS Version 8.3 contains the following enhancements for the PCSI utility:
The new PRODUCT ANALYZE PDB command verifies the structural integrity of the product database and, in some circumstances, performs minor repairs. This command:
For information on optional qualifiers for the command, see the
POLYCENTER Software Installation utility chapter in the HP OpenVMS System Management Utilities Reference Manual.
3.10.2 Automatic Verification of the Product Database
The PRODUCT INSTALL command has been enhanced to automatically verify
the product database at the start of the operation and again when the
product database is updated. The verification and repair actions are
similar to those provided by the PRODUCT ANALYZE PDB command. Other
PRODUCT commands that modify the database (such as PRODUCT RECONFIGURE
and PRODUCT REMOVE) also perform a verification pass.
3.10.3 Support for ODS-5 Volumes
The PCSI utility now provides full support for enhanced ODS-5 directory and file specification syntax. Prior to OpenVMS Version 8.3, product kits could be installed on ODS-5 disks, but all files were placed on the destination volume using ODS-2 syntax rules (for example, file names were in upper case only). With ODS-5 support, the kit developer can create a kit that installs files utilizing enhanced syntax such as:
The technique used to specify enhanced syntax is to quote the name specification parameter of the DIRECTORY or FILE statement in the product description file (PDF). For example, consider a PDF file that contains the following two lines:
file [test]file_one.txt ; file "[test]File_Two.txt" ; |
If the destination disk is an ODS-5 volume, the PCSI utility for OpenVMS Version 8.3 will install the first file as FILE_ONE.TXT and the second file as File_Two.txt. For compatibility with older versions of the PCSI utility, an unquoted specification is changed to all uppercase characters.
In addition, the product developer can test whether a volume is ODS-2 or ODS-5 by use of a new PDF statement of the form:
IF ( < FILESYSTEM { ODS-2 | ODS-5 } [ VOLUME { DESTINATION | SYSTEM } ] > ) ;
|
For example, to conditionally process statements if the destination volume is ODS-5, the following lines can be used:
IF ( < FILESYSTEM ods-5 VOLUME destination > ) ; ... END IF ; |
Several PRODUCT commands have been enhanced to support Secure Delivery of product kits. See Section 5.2 for an overview of Secure Delivery. The following capabilities have been added to the PCSI utility:
For more information about PRODUCT commands and qualifiers that support
secure delivery, see the POLYCENTER Software Installation utility
chapter in the HP OpenVMS System Management Utilities Reference Manual.
3.10.5 Defaults Changed on Two Qualifiers
In OpenVMS Version 8.3, default values for certain qualifiers are changed for the following commands:
The SANCP utility allows you to limit the number of active I/Os a host may have across all paths to Logical Unit Numbers (LUN) on a given Fibre Channel storage port. A storage port can be selected by a discrete or wildcarded port World Wide ID (WWID) or by a product ID substring.
The SANCP utility processes a command and qualifiers passed to it on
the command line, allowing it to be executed from a DCL script, or it
prompts you if started with no command. For more information about the
SANCP utility, see the SANCP Help facility.
3.12 SAS Utility (I64 Only)
The SAS utility (SAS$UTIL)is an OpenVMS system management and diagnostic tool that is capable of configuring Integrated RAID (IR) functionality for the HP 8 Internal Port Serial Attached SCSI Host Bus Adapter (SAS Controller).
Integrated RAID (IR) is used where extra performance, storage capacity, or redundancy of a RAID configuration, or all three, are required. OpenVMS Version 8.3 supports only Integrated RAID 1 or Integrated Mirroring (IM) and its associated Global Hot Spare capability.
For more detailed information, see the HP OpenVMS System Management Utilities Reference Manual.
3.13 SCACP utility
The following new features are added to the SCACP utility in OpenVMS Version 8.3:
These new features are described in the following sections.
3.13.1 Data Compression Management
The SCACP SET VC command now includes a /COMPRESSION (or /NOCOMPRESSION) qualifier, which enables or disables sending compressed data by the specified PEdriver VCs. The default is /NOCOMPRESSION.
You can also enable the VC use of compression by setting bit 3 of the
NISCS_PORT_SERV system parameter. The /NOCOMPRESSION qualifier does not
override compression enabled by setting bit 2 of NISCS_PORT_SERV.
3.13.2 Multi-Gigabit Scaling
You can use the SET VC COMMAND /WINDOW=RECEIVE_SIZE=value and /WINDOW=TRANSMIT_SIZE=value qualifiers to override the automatically calculated receive and transmit window sizes for a PEdriver VC. The /WINDOW=NORECEIVE_SIZE and /WINDOW=NOTRANSMIT_SIZE qualifiers can be used to remove management override of the window sizes. You can use /WINDOW=(NORECEIVE_SIZE,NOTRANSMIT_SIZE) to remove the override from both transmit and receive window sizes with a single command.
These new command qualifiers can be used to ensure that the VC has enough buffering to receive, and can transmit sufficient packets before waiting for return acknowledgments, to attain maximum bandwidth between nodes. To avoid unnecessary VC closures, these commands have restrictions on the order that they are issued. For information about these restrictions, see the HP OpenVMS System Management Utilities Reference Manual and the SCACP Help facility.
The new SCACP command, CALCULATE WINDOW_SIZE, can be used to determine the maximum window size that the VC should be using. This command has two required qualifiers, /DISTANCE=KILOMETERS (or MILES)=n and /SPEED=s, where n is the cable route distance between the two nodes, and s is the total bandwidth of all links being used for cluster communications between the nodes.
The SCACP SHOW VC command now displays if compression is enabled on a VC, and there are new columns for the management settings for transmit and receive window size.
For more information, see the SCACP utility chapter, and
NISCS_PORT_SERV in the HP OpenVMS System Management Utilities Reference Manual and HP OpenVMS Availability Manager User's Guide.
3.14 HP OpenVMS I64 Serial Multiplexer (MUX) Support (I64 Only)
RS232 serial lines and multiplexers are used for a variety of tasks, from traditional terminal connections to low-speed system-to-system communications and even communications with remote instruments. OpenVMS has traditionally supported adding serial lines at the same time as option-card-based multiplexers. This solution requires dedicating I/O slots; it also limits the choices of option cards available.
With the widespread adoption of the Universal Serial Bus (USB) on industry-standard platforms, OpenVMS has moved away from option-card-based multiplexers and has adopted USB to add serial lines to HP Integrity servers. Rather than using one or two option-card solutions with 8 or 16 lines for all configurations, you can now configure USB to meet your exact requirements.
Testing shows that the USB-based serial multiplexers perform as well as (or better than) their option-card counterparts and cause very low overhead to the system. In fact, the overhead is lower than option-card-based multiplexers.
For more information about HP MUX support, see the HP OpenVMS System Management Utilities Reference Manual.
3.15 Spinlock Trace Utility (SPL)
The Spinlock Trace utility available through SDA has been changed to report the various spinlock hold and wait times in nanoseconds instead of cycles. This allows for easy comparison of information collected among systems with different cycle counter frequencies.
The display has also been enhanced to show the full 64-bit address of
PCs as we are moving code to run in P2 and S2 space. Also the decoding
of the PC address has been enhanced to show the module, routine and
offset for I64. The SPL ANALYZE and SPL SHOW COLLECT commands displays
the additional PC decoding by default, but SPL SHOW TRACE [/SUMMARY]
shows the information only if the /FULL qualifier is specified because
the display is already congested.
3.16 HP OpenVMS System Analysis Tools
The following sections describe the new features provided in the System
Analysis Tools utilities. The HP OpenVMS System Analysis Tools Manual is not updated for this
release, but the additions and changes noted in this manual and in the
HP OpenVMS Version 8.3 Release Notes have been included in online help for the SDA utility and
in related commands for ANALYZE and System Service Logging.
3.16.1 System Dump Debugger
The System Dump Debugger (SDD) is now supported on OpenVMS for
Integrity servers as well as on OpenVMS Alpha.
3.16.2 System Dump Analyzer
The following sections describe the following new SDA or SDA extension commands and new callable routine extensions, as well as several new qualifiers for SDA commands:
The Common Bitmask Block (CBB) routines, SDA$CBB_xxx, are designed for use with local copies of the CBB structures that describe the CPUs in use in a system. The CBB structures are assumed to be at least CBB$K_STATIC_BLOCK bytes in length. The definitions of the various CBB constants and field names used by these routines can be found in CBBDEF.H in SYS$LIBRARY:SYS$LIB_C.TLB.
The set of routines is not intended to be an exhaustive set of all possible CBB-related operations, but provides those operations known to be needed. They may not work as expected with CBB structures set up for any purpose other than to describe CPUs.
Collects file identification to file name translation data on both OpenVMS Alpha and OpenVMS for Integrity servers, and processes unwind data only on OpenVMS for Integrity servers.
COLLECT [qualifiers]
None
/LOG
Displays information on the progress of the COLLECT command, for example, the name of the process being scanned, or (on Integrity servers) the name of an image whose unwind data is being collected./SAVE [= file-name]
Writes collection data to a separate file. By default, a file of type .COLLECT with the same name as the dump file is created in the same directory as the dump file./UNDO
Removes all the file or unwind data from an earlier COLLECT command from SDA's memory. COLLECT/UNDO does not affect the file or unwind data already appended to the dump file being analyzed, or already written to a separate collection file.
When a dump is being analyzed, it is useful to have data available that cannot be written to the dump file at the time of the system crash. This data includes the full file specification associated with a file identification. On OpenVMS for Integrity servers, it also contains the unwind data for images activated in processes.If the dump is being analyzed on the system where it is originally written, this data can be collected for use in the current SDA session using the COLLECT command. If the dump is being copied for analysis elsewhere, the COPY/COLLECT command can be used to collect the data and append it to the copy being written. If the COPY/COLLECT command is used after a COLLECT command, the data already collected is appended to the dump copy.
For all file or unwind data to be collected successfully, all disks that were mounted at the time of the system crash should be remounted and accessible to the process running SDA.
If the COPY and the COLLECT cannot be done as a single step, a COLLECT/SAVE command writes the collection to a separate file that can be used later with the dump file. A subsequent COPY command combines the two files.
Displays information about scheduling classes that are active in the system or dump being analyzed.
SHOW CLASS [class-name | /ALL]
class-name
Name of the class to be displayed.
/ALL
Indicates that details of all active classes are to be displayed.
SDA displays information about active scheduling classes in the system. By default, a summary of the classes is displayed.
Displays information from the Extensible Firmware Interface (EFI) data structures. Currently, the only display provided by SDA is the EFI memory map.
SHOW EFI /MEMMAP [=ALL] [range]
range
The entry or range of entries to be displayed, expressed using the following syntax:m---Displays the entry m.
m:n---Displays the entries from m to n.
m;n---Displays n entries starting at m.You cannot specify a range with /MEMMAP=ALL.
/MEMMAP [=ALL]
Displays the EFI memory map. This qualifier is required. By default, only entries in the EFI memory map with the runtime attribute are displayed. If the /MEMMAP=ALL qualifier is specified, all entries are displayed.You cannot specify the /MEMMAP=ALL qualifier and supply a range of entries to be displayed.
SDA locates the EFI memory map in the system or dump and displays the contents. If no range is given, SDA also displays information about the location and size of the memory map.
Displays data from the Virtual Hash Page Table.
SHOW VHPT [ /CPU = { n | * } [ /ALL ] [ range ] ]
range
The entry or range of entries to be displayed, expressed using the following syntax:m---Displays the VHPT entry m.
m:n---Displays the VHPT entries from m to n.
m;n---Displays n VHPT entries starting at m.A range can only be provided if a single CPU is specified with the /CPU qualifier.
/CPU = { n | * }
Indicates that the detailed contents of the VHPT for one or all CPUs is to be displayed. The default action is for a summary of VHPT information to be displayed./ALL
Displays all VHPTs for the specified CPU(s). Without /ALL, only entries that have a valid tag are displayed.
Displays contents of the Virtual Hash Page Table on an OpenVMS I64 system. By default, a summary of the VHPT entries is displayed. If CPUs are specified, details of individual VHPT entries are displayed for the CPUs. If a single CPU is specified, specific VHPT entries for that CPU are displayed.In the detailed display, the columns are as follows:
Table 3-3 Entry VHPT Entry Number Bits One or more of the following flags: P---Present
A---Accessed
D---Dirty
E---Exception deferral
I---Tag invalid (only seen if /ALL is specified)MA One of the following memory attributes: WB---Write Back
UC---Uncacheable
UCE---Uncacheable Exported
WC---Write Coalescing
NaT---NaTPageAR/PL The access rights and privilege level of the page. Consists of a number (0-7) and a letter (K, E, S, or U) that determines access to the page in each mode. KESU The access allowed to the page in each mode. This is an interpretation of the AR/PL values in the previous column. For an explanation of the access codes, see HP OpenVMS System Analysis Tools Manual. Physical address The starting physical address for this VHPT entry. Page size The size of the page represented by this VHPT entry. Page sizes for VHPT entries range from 4KB to 4GB. Not all possible pages sizes are used by OpenVMS for Integrity servers. Tag The translation tag for the VHPT entry. Quad4 Information recorded by OpenVMS for Integrity servers for debugging purposes. The contents of this quadword are subject to change.
| Previous | Next | Contents | Index |