HP OpenVMS System Management Utilities Reference Manual

Contents

Index

Table 19-5 COUNTERS Class Fields
Field Name Description

BDT_WAITS Number of times this connection had to wait for a buffer descriptor.

BLKS_REQ Number of block-request data commands initiated to block transfer data from the remote system to the local system.

BLKS_SENT Number of block-send data commands initiated to block-transfer data from the local system to the remote system.

CR_WAITS Number of times this connection had to wait for send credit.

DGS_DSCRD Number of application datagrams discarded by the port driver.

DGS_RCVD Number of application datagrams received by the local system over the connection from the remote system and given to SYSAP.

DGS_SENT Number of application datagrams sent over the connection.

KB_MAPPED Number of kilobytes of data mapped for block transfer.

KB_RCVD Number of kilobytes of data received by the local system from the remote system through request-data commands.

KB_SENT Number of kilobytes of data sent from the local system to the remote system through send-data commands.

MSGS_RCVD Number of application datagram messages received over the connection.

MSGS_SENT Number of application datagram messages sent over the connection.

**Table 19-5 COUNTERS Class Fields**
Field Name	Description
BDT_WAITS	Number of times this connection had to wait for a buffer descriptor.
BLKS_REQ	Number of block-request data commands initiated to block transfer data from the remote system to the local system.
BLKS_SENT	Number of block-send data commands initiated to block-transfer data from the local system to the remote system.
CR_WAITS	Number of times this connection had to wait for send credit.
DGS_DSCRD	Number of application datagrams discarded by the port driver.
DGS_RCVD	Number of application datagrams received by the local system over the connection from the remote system and given to SYSAP.
DGS_SENT	Number of application datagrams sent over the connection.
KB_MAPPED	Number of kilobytes of data mapped for block transfer.
KB_RCVD	Number of kilobytes of data received by the local system from the remote system through request-data commands.
KB_SENT	Number of kilobytes of data sent from the local system to the remote system through send-data commands.
MSGS_RCVD	Number of application datagram messages received over the connection.
MSGS_SENT	Number of application datagram messages sent over the connection.

Table 19-6 CREDITS Class Fields
Field Name Description

INIT_REC Initial receive credit extended to the remote system when the connection was made.

MIN_REC Minimum receive credit (minimum send credit required by the remote system).

MIN_SEND Minimum send credit.

PEND_REC Receive credit not yet extended to the remote system.

RECEIVE Receive credit (send credit held by the remote system).

SEND Current send credit.

**Table 19-6 CREDITS Class Fields**
Field Name	Description
INIT_REC	Initial receive credit extended to the remote system when the connection was made.
MIN_REC	Minimum receive credit (minimum send credit required by the remote system).
MIN_SEND	Minimum send credit.
PEND_REC	Receive credit not yet extended to the remote system.
RECEIVE	Receive credit (send credit held by the remote system).
SEND	Current send credit.

Table 19-7 ERRORS Class Fields
Field Name Description

ERT_COUNT Number of port reinitialization attempts remaining.

ERT_MAX Total number of times a recovery from fatal port errors can be attempted by shutting down all virtual circuits and connections and reinitializing the port.

NUM_ERRORS Number of errors that have been logged on the port since the system was booted. This number includes errors encountered in reinitialization attempts as well as recoverable errors, such as virtual circuit failure. This is the same error count as that displayed by the DCL command SHOW DEVICE.

**Table 19-7 ERRORS Class Fields**
Field Name	Description
ERT_COUNT	Number of port reinitialization attempts remaining.
ERT_MAX	Total number of times a recovery from fatal port errors can be attempted by shutting down all virtual circuits and connections and reinitializing the port.
NUM_ERRORS	Number of errors that have been logged on the port since the system was booted. This number includes errors encountered in reinitialization attempts as well as recoverable errors, such as virtual circuit failure. This is the same error count as that displayed by the DCL command SHOW DEVICE.

Table 19-8 LOCAL_PORTS Class Fields
Field Name Description

BUFF_DESCR Number of buffer descriptors in use.

CMDS_QUEUED Total number of messages, datagrams, and port commands queued for transmission at all priorities by the port.

COUNTER_OWNER Name of the process currently using the port traffic counters.

DGI_MAP A 16-bit bit map displayed as four hexadecimal digits. Each bit in the map represents a port in the cluster from which datagram reception has been disabled.

DG_OVRHD_SIZE Number of bytes of port header, SCS header, and DECnet header in a datagram.

DGS_FREE Number of free datagram buffers currently queued for receive commands.

FORM_CIRCS Number of formative circuits (circuits in the process of opening) from the port.

FREE_BUFF Number of CI buffer descriptors free for use.

LB_STATUS Loopback status of each cable from the port to the star coupler. The field applies only to CI. Possible displays are as follows:

A-B Loopback tests pass on paths A and B.

A-# Loopback tests pass on path A.

#-B Loopback tests pass on path B.

#-# Loopback tests failed on paths A and B.

N/A Loopback tests are not being done.

LOG_MAP A 16-bit bit map displayed as four hexadecimal digits. Each bit in the map represents a port in the cluster for which an error was logged.
Errors are logged when data provided by the configuration database on the local system conflicts with data provided by the remote system. When a conflict is discovered and an error is logged, virtual circuits to the remote system can no longer be established.

LP_LD_CLASS Hard-coded capacity value of the port, based on the megabits/second rate of the interconnect of the port.

LP_PRIORITY Management priority assigned to the port.

LP_STATUS Status of the local port. The port is either on line or off line.

LP_TYPE Device type of the port (CI780, CI750).

MAX_PORT Largest port number to which a virtual circuit open is attempted.

MSGS_FREE Number of free message buffers currently queued for receives commands.

MSG_HDR_SIZE Number of bytes of port header and SCS header in a message.

NAME Device name of the local port.

OPEN_CIRCS Number of virtual circuits open from the port.

POOL_WAITERS Number of processes waiting for nonpaged pool resources for message buffers.

PORT_NUM Port number assigned to the port.

PRT_MAP A 16-bit bit map displayed as three hexadecimal digits. Each bit in the map represents a port in the cluster that has been recognized by the host system.

RSPS_QUEUED Total number of responses of all kinds received from the port but not yet processed.

Table 19-9 MEMBERS Class Fields
Field Name Description

ACK_LM Maximum number of OpenVMS Cluster messages the remote system can receive before sending an acknowledgment.

ACKR_SQ Sequence number of the last acknowledgment received over the cluster connection.

CNX_STATE State of the cluster connection. Possible displays are as follows:

ACCEPT Initial connection is accepted.

CLOSED Connection is closed.

CONNECT Initial connection is being accepted.

DEAD No connection is possible.

DISCONNECT Disconnection is in progress.

NEW No attempt to make a connection has been made yet.

OPEN Connection is open.

REACCEPT Connection is accepting the reconnect request.

RECONNECT Connection is attempting to reconnect.

WAIT Timeout is in progress.

CSID Cluster system identification number. This number is unique over the life of the cluster. Unlike SYS_ID, this identification number may change when the system reboots.

DIR_WT Lock manager distributed directory weight.

EXPECTED_VOTES Maximum number of votes that an individual node can encounter. Used as an initial estimate for computing CL_EXPECTED_VOTES.
The cluster manager sets this number using the EXPECTED_VOTES system parameter. It is possible for this field to display a number smaller than the EXPECTED_VOTES parameter setting if the REMOVE_NODE option was used to shut down a cluster member or the SET CLUSTER/EXPECTED_VOTES DCL command was used since this node was last rebooted.
The dynamic value for EXPECTED_VOTES used clusterwide is the CL_EXPECTED_VOTES field, which is described in Table 19-3.

PROTOCOL Protocol version number and ECO level of the connection management software.

QDVOTES Number of votes the remote system recommends be contributed by the quorum disk. Normally, the cluster manager sets this number using the system parameter QDSKVOTES.

QF_ACTIVE Indicates whether the remote system's quorum file is accessible.

QF_SAME Indicates whether the local and remote systems agree about which disk is the quorum disk.

QF_WATCHER Remote system has an active connection to the quorum disk and can verify its connection for members unable to access the disk directly.

QUORUM Derived from EXPECTED_VOTES and calculated by the connection manager. It represents an initial value for the minimum number of votes that must be present for this node to function. The dynamic QUORUM value is the CL_QUORUM field, which is described in the CLUSTER class category in Table 19-3.

RCVD_SQ Sequence number of the last message received over the OpenVMS Cluster connection.

RECNXINTERVAL Displays the time (in seconds) that the connection manager will wait before timing out the corresponding connection. It is the maximum of the value contained in the system parameter RECNXINTERVAL on the local node and the amount of time it would take for the connection manager on the remote node to discover that the connection is broken.

SEND_SQ Sequence number of the next message to be sent over the OpenVMS Cluster connection.

STATUS Status of the node in the cluster. Possible displays are as follows:

blank System is not being considered as a cluster member.

BRK_MEM System is a member of the cluster, but the connection manager has lost communication with it.

BRK_NEW System has just booted, but has not yet joined the cluster and the connection manager has lost communication with it.

BRK_NON Connection manager has lost communication with the system and the system is no longer a member of the cluster.

BRK_REM Connection manager has lost communication with the system, and the system has been removed from the cluster.

MEMBER System is participating in the cluster.

NEW System has just booted, but has not yet joined the cluster. If this system would normally be a member of the cluster and is displaying NEW in this field, you can expect that the display will eventually change to MEMBER.

NON System is not a member of the cluster.

REMOVED System has been removed from the cluster.

SW_VERS Indicator of the software version running on the node.

TRANSITION_TIME Time of the system's last change in membership status. (See the STATUS field.)

UNACKED Number of unacknowledged OpenVMS Cluster messages received by the remote system.

VOTES Number of votes the remote node contributes toward quorum. Normally, the cluster manager sets this number with the system parameter VOTES.

WARMCDRPS Number of CDRPs on the CDRP free queue.

Table 19-10 SYSTEMS Class Fields
Field Name Description

DG_SIZE Maximum number of bytes of application data in datagrams sent over the circuit.

HW_TYPE System hardware type (for example, VAXstation 3100 or HS70). (Enclose the system type between double quotation marks.)

HW_VERS Hardware configuration and revision levels of the remote system.

INCARNATION Unique 16-digit hexadecimal number established when the system is booted.

INCN_TIME Incarnation number expressed as a time (dd-mmm-yy hh:mm).

MSG_SIZE Maximum number of bytes of application data in messages sent over the circuit.

NODE Node name of the remote system. Normally, the cluster manager sets the node name using the system parameter SCSNODE. The node name should be the same as the DECnet node name. (Note that SCSNODE cannot be more than six characters.)

NUM_CIRCUITS Number of virtual circuits between the local system and remote systems.

SOFTWARE Name and version of the operating system currently running on the remote system.

SYS_ID Identification number of the remote system. Normally, the cluster manager sets this number using the system parameters SCSSYSTEMID and SCSSYSTEMIDH. This number should be the same as the DECnet node number.

**Table 19-10 SYSTEMS Class Fields**
Field Name	Description
DG_SIZE	Maximum number of bytes of application data in datagrams sent over the circuit.
HW_TYPE	System hardware type (for example, VAXstation 3100 or HS70). (Enclose the system type between double quotation marks.)
HW_VERS	Hardware configuration and revision levels of the remote system.
INCARNATION	Unique 16-digit hexadecimal number established when the system is booted.
INCN_TIME	Incarnation number expressed as a time (dd-mmm-yy hh:mm).
MSG_SIZE	Maximum number of bytes of application data in messages sent over the circuit.
NODE	Node name of the remote system. Normally, the cluster manager sets the node name using the system parameter SCSNODE. The node name should be the same as the DECnet node name. (Note that SCSNODE cannot be more than six characters.)
NUM_CIRCUITS	Number of virtual circuits between the local system and remote systems.
SOFTWARE	Name and version of the operating system currently running on the remote system.
SYS_ID	Identification number of the remote system. Normally, the cluster manager sets this number using the system parameters SCSSYSTEMID and SCSSYSTEMIDH. This number should be the same as the DECnet node number.

You can customize the SHOW CLUSTER display to include the information most important to your needs by dynamically adding and removing classes and fields. For example, if you add a field that belongs to the CLUSTER class or the LOCAL_PORTS class, SHOW CLUSTER adds the new column of information to the display.

By default, the SHOW CLUSTER display includes the NODE and SOFTWARE fields of the SYSTEMS class and the STATUS field of the MEMBERS class. Figure 19-1 presents a sample customized SHOW CLUSTER display in which the HW_TYPE, VOTES, and TRANSITION_TIME fields have been added to the default SHOW CLUSTER display.

Figure 19-1 Sample SHOW CLUSTER Report

Because SHOW CLUSTER information covers approximately 100 fields of data, the display can quickly extend beyond screen limits. Therefore, the utility provides mechanisms to help you control the display of data. These mechanisms include the following ones:

SHOW CLUSTER qualifiers
SHOW CLUSTER commands
A default keypad, which can be redefined
An initialization file to format the display
Command procedures to control the display

SHOW CLUSTER has a number of qualifiers and commands, and a definable keypad that allow you to customize the display. You can rearrange the position of windows, scroll their contents, or change the interval at which the display is updated. This chapter contains reference information for the SHOW CLUSTER qualifiers and commands. Appendix I describes how to use the keypad.

Over time, as you determine the most valuable classes and fields of data for the SHOW CLUSTER report, you can create a startup initialization file that establishes your default report format. You can also build command procedures and define a keypad to use while running SHOW CLUSTER interactively.

By customizing SHOW CLUSTER output, you can display only data that is relevant to your installation. Creating the initialization file SHOW_CLUSTER$INIT:SHOW_CLUSTER.INI is described in HP OpenVMS System Manager's Manual, Volume 2: Tuning, Monitoring, and Complex Systems.

By customizing the SHOW CLUSTER keypad, you can redefine default keypad functions to be more site specific. Using SHOW CLUSTER keypad commands is described in Appendix I.

19.2 SHOW CLUSTER Usage Summary

The Show Cluster utility (SHOW CLUSTER) monitors the activity and performance of an OpenVMS Cluster system, and outputs the information to your default terminal or to a specified device or file.

Format

SHOW CLUSTER [/qualifier[,...]]

Parameters

None.

Description

To invoke SHOW CLUSTER, enter the following command:
$ SHOW CLUSTER
If you specify the command without any qualifiers, SHOW CLUSTER displays a single cluster report and then returns control to the DCL level. To invoke a continuous SHOW CLUSTER display, enter the following command:
$ SHOW CLUSTER/CONTINUOUS
In a continuous display, you can control report output with SHOW CLUSTER commands. You can direct SHOW CLUSTER output to a file or device other than to SYS$OUTPUT by specifying the /OUTPUT qualifier with the SHOW CLUSTER command.
To exit from a continuous display and return to the DCL level, enter the EXIT command or press Ctrl/Z. To exit from SHOW CLUSTER without erasing the screen, press Ctrl/C. To interrupt SHOW CLUSTER, press Ctrl/Y.

Contents

Index