Document revision date: 19 July 1999
[Compaq] [Go to the documentation home page] [How to order documentation] [Help on this site] [How to contact us]
[OpenVMS documentation]

OpenVMS System Manager's Manual


Previous Contents Index

20.6 Using the SYSMAN Command DO to Manage an OpenVMS Cluster

The SYSMAN command DO enables you to execute a DCL command or command procedure on all nodes in the current environment. This is convenient when you are performing routine system management tasks on nodes in the OpenVMS Cluster system, such as:

Each DO command executes as an independent process, so there is no process context retained between DO commands. For this reason, you must express all DCL commands in a single command string, and you cannot run a program that expects input.

In a cluster environment, SYSMAN executes the commands sequentially on all nodes in the cluster. Each command executes completely before SYSMAN sends it to the next node in the environment. Any node that is unable to execute the command returns an error message. SYSMAN displays an error message if the timeout period expires before the node responds.

In a dual-architecture heterogeneous OpenVMS Cluster running both OpenVMS VAX and OpenVMS Alpha, some uses of the DO command may require special handling. For example, if you are installing images that are named differently in each architecture, you can still use the DO command if you create logical name tables for VAX and for Alpha nodes. See the example sequence that follows this description for an example.

Some DCL commands, such as MOUNT/CLUSTER or SET QUORUM/CLUSTER, operate clusterwide by design. Similarly, operations on clusterwide logical names and tables operate clusterwide by design. It is best to avoid using these kinds of commands with the DO command in SYSMAN when the environment is set to cluster. As alternatives, you could leave SYSMAN temporarily with the SPAWN command and execute these commands in DCL, or you could define the environment to be a single node within the cluster.

Examples

  1. The following example installs an image on a cluster. First, it adds CMKRNL and SYSPRV privileges to the current privileges because they are required by INSTALL and AUTHORIZE. The DO INSTALL command installs the file STATSHR. The DO MCR AUTHORIZE command sets up an account for user Jones, specifying a password and a default device and directory.


    SYSMAN> SET PROFILE/PRIVILEGES=(CMKRNL,SYSPRV)/DEFAULT=SYS$SYSTEM
    SYSMAN> DO INSTALL ADD/OPEN/SHARED WRKD$:[MAIN]STATSHR
    SYSMAN> DO MCR AUTHORIZE ADD JONES/PASSWORD=COLUMBINE -
    _SYSMAN> /DEVICE=WORK1/DIRECTORY=[JONES]
    

  2. The following example sets the environment to cluster and starts up a software product called XYZ on each node in the cluster:


    SYSMAN>SET ENVIRONMENT/CLUSTER
    %SYSMAN-I-ENV, Current command environment:
            Clusterwide on local cluster 
            Username SMITH    will be used on nonlocal nodes
    SYSMAN> DO @SYS$STARTUP:XYZ_STARTUP
    

  3. The following example shows how you can define logical names for VAX and Alpha nodes in a dual-architecture heterogeneous cluster, so that you can use the DO command to install architecture-specific images.


    $ CREATE/NAME_TABLE/PARENT=LNM$SYSTEM_DIRECTORY SYSMAN$NODE_TABLE
    $ DEFINE/TABLE=SYSMAN$NODE_TABLE ALPHA_NODES NODE21,NODE22,NODE23    
    $ DEFINE/TABLE=SYSMAN$NODE_TABLE VAX_NODES NODE24,NODE25,NODE26    
    $ RUN SYS$SYSTEM:SYSMAN
    SYSMAN> SET ENVIRONMENT/NODE=ALPHA_NODES
    %SYSMAN-I-ENV, current command environment: 
             Individual nodes: NODE21,NODE22,NODE23 
             Username BOUCHARD will be used on nonlocal nodes
     
    SYSMAN> DO INSTALL REPLACE SYS$LIBRARY:DCLTABLES.EXE
    %SYSMAN-I-OUTPUT, command execution on node NODE21 
    %SYSMAN-I-OUTPUT, command execution on node NODE22 
    %SYSMAN-I-OUTPUT, command execution on node NODE23
    SYSMAN> DO INSTALL REPLACE SYS$SYSTEM: DEC_FORTRAN.EXE
    %SYSMAN-I-OUTPUT, command execution on node NODE21 
    %SYSMAN-I-OUTPUT, command execution on node NODE22 
    %SYSMAN-I-OUTPUT, command execution on node NODE23
     
    SYSMAN> SET ENVIRONMENT/NODE=VAX_NODES
    %SYSMAN-I-ENV, current command environment: 
             Individual nodes: NODE24,NODE25,NODE26 
             Username BOUCHARD will be used on nonlocal nodes
     
    SYSMAN> DO INSTALL REPLACE SYS$LIBRARY:DCLTABLES.EXE
    %SYSMAN-I-OUTPUT, command execution on node NODE24 
    %SYSMAN-I-OUTPUT, command execution on node NODE25 
    %SYSMAN-I-OUTPUT, command execution on node NODE26
    SYSMAN> DO INSTALL REPLACE SYS$SYSTEM:FORTRAN$MAIN.EXE
    %SYSMAN-I-OUTPUT, command execution on node NODE24 
    %SYSMAN-I-OUTPUT, command execution on node NODE25 
    %SYSMAN-I-OUTPUT, command execution on node NODE26
    

  4. The following example shows which files are open on DISK2. You might use this if you want to dismount DISK2 and need to see which users in the cluster have files open.


    SYSMAN >SET ENVIRONMENT/CLUSTER
    %SYSMAN-I-ENV, Current command environment:
            Clusterwide on local cluster 
            Username SMITH    will be used on nonlocal nodes
    SYSMAN> DO SHOW DEVICE/FILES DISK2:
     
    %SYSMAN-I-OUTPUT, command execution on node NODE21 
    Files accessed on device $1$DIA2: (DISK2, NODE22) on 14-MAY-1998 15:44:06.05 
    Process name      PID     File name 
                    00000000  [000000]INDEXF.SYS;1 
    %SYSMAN-I-OUTPUT, command execution on node NODE22 
    Files accessed on device $1$DIA2: (DISK2, NODE21) on 14-MAY-1998 15:44:26.93 
    Process name      PID     File name 
                    00000000  [000000]INDEXF.SYS;1 
    %SYSMAN-I-OUTPUT, command execution on node NODE23 
    Files accessed on device $1$DIA2: (NODE21, NODE22) on 14-MAY-1998 15:45:01.43 
    Process name      PID     File name 
                    00000000  [000000]INDEXF.SYS;1 
    %SYSMAN-I-OUTPUT, command execution on node NODE24 
    Files accessed on device $1$DIA2: (NODE22, NODE21) on 14-MAY-1998 15:44:31.30 
    Process name      PID     File name 
                    00000000  [000000]INDEXF.SYS;1 
    Susan Scott     21400059  [SCOTT]DECW$SM.LOG;228 
    _FTA7:          214000DD  [SCOTT]CARE_SDML.TPU$JOURNAL;1 
    %SYSMAN-I-OUTPUT, command execution on node NODE25 
    Files accessed on device $1$DIA2: (NODE21, NODE22) on 14-MAY-1998 15:44:35.50 
    Process name      PID     File name 
                    00000000  [000000]INDEXF.SYS;1 
    DECW$SESSION    226000E6  [SNOW]DECW$SM.LOG;6 
    _FTA17:         2260009C  [SNOW.MAIL]MAIL.MAI;1 
    SNOW_1          2260012F  [SNOW.MAIL]MAIL.MAI;1 
    SNOW_2          22600142  [SNOW.MAIL]MAIL.MAI;1 
    SNOW_3          22600143  [SNOW.MAIL]MAIL.MAI;1 
    

  5. The following example shows how much memory is available on the nodes in a cluster. You might use this if you are installing software and want to know if each node has enough memory available.


    SYSMAN > SET ENVIRONMENT/NODE=(NODE21,NODE22)
    %SYSMAN-I-ENV, Current command environment:
            Clusterwide on local cluster 
            Username SMITH    will be used on nonlocal nodes
    SYSMAN>  DO SHOW MEMORY
    %SYSMAN-I-OUTPUT, command execution on node NODE21 
                  System Memory Resources on 14-MAY-1998 15:59:21.61 
    Physical Memory Usage (pages):     Total        Free      In Use    Modified 
      Main Memory (64.00Mb)           131072       63955       65201        1916 
    Slot Usage (slots):                Total        Free    Resident     Swapped 
      Process Entry Slots                360         296          64           0 
      Balance Set Slots                  324         262          62           0 
    Fixed-Size Pool Areas (packets):   Total        Free      In Use        Size 
      Small Packet (SRP) List          10568        1703        8865         128 
      I/O Request Packet (IRP) List     3752         925        2827         176 
      Large Packet (LRP) List            157          28         129        1856 
    Dynamic Memory Usage (bytes):      Total        Free      In Use     Largest 
      Nonpaged Dynamic Memory        1300480       97120     1203360       60112 
      Paged Dynamic Memory           1524736      510496     1014240      505408 
    Paging File Usage (pages):                      Free  Reservable       Total 
      DISK$MTWAIN_SYS:[SYS0.SYSEXE]SWAPFILE.SYS                                     
                                                   10000       10000       10000 
      DISK$MTWAIN_SYS:[SYS0.SYSEXE]PAGEFILE.SYS                                     
                                                   60502      -52278      100000 
    Of the physical pages in use, 19018 pages are permanently allocated to VMS. 
     
    %SYSMAN-I-OUTPUT, command execution on node NODE22 
                  System Memory Resources on 14-MAY-1998 15:59:42.65 
    Physical Memory Usage (pages):     Total        Free      In Use    Modified 
      Main Memory (32.00Mb)            65536       44409       20461         666 
    Slot Usage (slots):                Total        Free    Resident     Swapped 
      Process Entry Slots                240         216          24           0 
      Balance Set Slots                  212         190          22           0 
    Fixed-Size Pool Areas (packets):   Total        Free      In Use        Size 
      Small Packet (SRP) List           5080        2610        2470         128 
      I/O Request Packet (IRP) List     3101        1263        1838         176 
      Large Packet (LRP) List             87          60          27        1856 
    Dynamic Memory Usage (bytes):      Total        Free      In Use     Largest 
      Nonpaged Dynamic Memory        1165312      156256     1009056      114432 
      Paged Dynamic Memory           1068032      357424      710608      352368 
    Paging File Usage (pages):                      Free  Reservable       Total 
      DISK$MTWAIN_SYS:[SYS1.SYSEXE]SWAPFILE.SYS                                     
                                                   10000       10000       10000 
      DISK$MTWAIN_SYS:[SYS1.SYSEXE]PAGEFILE.SYS                                     
                                                  110591       68443      120000 
    Of the physical pages in use, 9056 pages are permanently allocated to VMS. 
    


Chapter 21
Network Considerations

This chapter introduces the basic network software options available for OpenVMS Systems. Material provided in this chapter is intended as an introduction only: refer to the appropriate documentation set for the network product or products you are using for complete planning, installation, configuration, use, and management information.

21.1 Network Options Available on OpenVMS Systems

On OpenVMS systems, three types of network functionality are available:

Nodes running DECnet-Plus, TCP/IP, and DECnet Phase IV can coexist in the same network. You can run TCP/IP software and either DECnet-Plus or DECnet Phase IV on the same system. Table 21-1 lists the various software combinations possible on a node and which applications can be used for communication between various pairs of systems.

Table 21-1 Network Software Interoperability Options
If System A
Has...
And System B
Has...
Then Systems A and B
Can Communicate Using ...
TCP/IP TCP/IP TCP/IP applications
DECnet Phase IV DECnet Phase IV DECnet applications
DECnet-Plus DECnet-Plus DECnet applications
OSI applications
DECnet-Plus DECnet Phase IV DECnet applications
DECnet-Plus OSI OSI applications
TCP/IP and DECnet Phase IV TCP/IP TCP/IP applications
TCP/IP and DECnet Phase IV DECnet Phase IV DECnet applications
TCP/IP and DECnet-Plus TCP/IP TCP/IP applications
TCP/IP and DECnet-Plus DECnet-Plus DECnet applications
OSI applications
TCP/IP and DECnet-Plus TCP/IP and DECnet-Plus OSI applications
DECnet applications
DECnet applications via
DECnet over TCP/IP (RFC 1859)+
OSI applications via
OSI over TCP/IP (RFC 1006)
TCP/IP applications
TCP/IP and DECnet-Plus OSI (supporting RFC 1006) and TCP/IP OSI applications
OSI over TCP/IP (RFC 1006)
TCP/IP applications
TCP/IP and DECnet-Plus OSI (not supporting RFC 1006) and TCP/IP OSI applications
TCP/IP applications


+RFC 1859 is an Internet draft, an extension of Internet standard RFC 1006.

For an introduction to DECnet-Plus and a roadmap of the documentation set, refer to DECnet-Plus for OpenVMS Introduction and User's Guide.

For an introduction to DIGITAL TCP/IP Services for OpenVMS, refer to the DIGITAL TCP/IP Services for OpenVMS Concepts and Planning Guide.

A comprehensive list of DECnet-Plus and TCP/IP Services for OpenVMS documentation is provided at the end of this chapter (see Section 21.4.)

The following sections introduce DECnet-Plus and DIGITAL TCP/IP Services for OpenVMS.

21.2 Understanding DECnet-Plus for OpenVMS Networks

DECnet-Plus for OpenVMS provides the means for various Compaq operating systems to communicate with each other and with systems provided by other vendors. The DECnet-Plus network supports remote system communication, resource sharing, and distributed processing. Network users can access resources on any system in the network. Each system participating in the network is known as a network node. In addition, DECnet-Plus includes support for the Internet standard RFC 1006 and the Internet draft RFC 1859, allowing OSI and DECnet applications to run over TCP/IP. Thus using DECnet-Plus, applications can connect to and communicate with peer OSI and DECnet applications on any DECnet Phase IV-based system or OSI-based system, whether from Compaq or from other vendors.

Table 21-2 defines terms related to DECnet-Plus networks.

Table 21-2 DECnet-Plus for OpenVMS Network Terminology
Term Definition
Address/Address tower DECnet-Plus systems have multiple address towers, also called protocol stacks, that describe various sets of communications protocols available for a particular node. These towers are stored in the namespace. They are used for determining the protocols that two nodes have in common so that they can communicate with each other.
Autoconfigure An option supported by DECnet-Plus in which you can have your end node's network entity title (NET) automatically configured by the adjacent router.
Domain Collection of systems that use the same routing protocol
Entity An individual, manageable piece of a network that has attributes that describe it, a name that identifies it, and an interface that supports management operations. Examples of entities are node, routing, and OSI transport.
Extended address A DECnet-Plus network address that does not fall within the limits of DECnet Phase IV addressing and thereby provides extended addressing capabilities. A DECnet-Plus network address can also be DECnet Phase IV compatible. The DECnet-Plus configuration procedure automatically builds an extended address from the Phase IV address of your node. The extended address should be of concern only if users and applications require extended addressing for communication with other OSI systems (Compaq or non-Compaq).
Multihome The ability to assign more than one network address to a system. Having multiple addresses allows you to have both a DECnet-Plus extended address, a Phase IV compatible address, and a TCP/IP address, so you can communicate with DECnet Phase IV, OSI (or DECnet-Plus), and TCP/IP systems. This also allows you to belong to more than one network.
Name service Software that manages node name and addressing information. DECnet-Plus offers a choice of three, distinct name services: Local namespace, Digital Distributed Name Service (DECdns), and the Domain Name System (DNS/BIND).
Namespace The set of names stored by, and accessible to, a name service.
Network entity title In OSI terminology, a network entity title (NET) is a network address that is used for identifying the Network layer protocol for routing. A DECnet-Plus system can automatically construct (autoconfigure) a NET for each transport operating over routing.
Network service access point (NSAP) One of the following addresses:
  • The global network address of a DECnet-Plus system
  • The addressable point at which a network entity provides the network service to a network user
  • The complete address that identifies both the particular network system and the transport module on that system that is to receive the data
The NSAP is used to determine the destination node for all packets and so must be unique in the network. The NSAP is a NET with a selector field other than 00. (A selector field identifies the transport to be used.)
Object/Application In DECnet Phase IV, an object is a process to which a logical link connects. Objects are set up by layered products that use DECnet. Some objects are DECnet system programs---for example, the Mail object; other objects are user-written programs.

In DECnet-Plus, objects are referred to as applications. Where Phase IV has an object database, DECnet-Plus has an applications database.

Phase IV compatible address A DECnet-Plus network address that falls within the limits of Phase IV addressing; that is, conforming to the Phase IV area and node limits, where the area number is from 1 to 63, and the node number is from 1 to 1023, as in 36.515. Your DECnet-Plus system needs a Phase IV compatible address to communicate with DECnet Phase IV nodes in the same network.
Time service Software that synchronizes the system clocks in computers connected by a network. The Digital Distributed Time Service (DECdts) enables distributed applications to execute in the proper sequence even though they run on different systems.

21.2.1 DECnet-Plus Features

DECnet-Plus provides many features designed to enhance networking capabilities. These features include:

To fully benefit from these new features, you may need to alter your present network. Section 21.2.4 reviews the decisions you must make prior to upgrading.


Previous Next Contents Index

  [Go to the documentation home page] [How to order documentation] [Help on this site] [How to contact us]  
  privacy and legal statement  
6017PRO_088.HTML