Job Management Partner 1/Software Distribution Client Description and User's Guide

[Contents][Glossary][Index][Back][Next]

6.4.2 Syntax of the relay system settings file

This section explains the content to be specified in the relay system settings file.

Organization of this subsection
(1) Format
(2) Details

(1) Format

ManagingHost host-name
WorkstationType MASTER
ProtocolKind TCP
[ConnectionKind {HOST|MASTER}]
[ClientActionProtocol {UDP|TCP}]
[ClientActionRetryCount retry-count]
[ClientActionRetryInterval retry-interval]
[MaxConnectClients number-of-clients]
[WatchTimeofStatus monitoring-interval]
[WatchTimeofOrders monitoring-interval]
[RetryCount retry-count]
[RetryInterval retry-interval]
[ServerRetryCount retry-count]
[ServerRetryInterval retry-interval]
[DeliveryRetryCount retry-count]
[DeliveryRetryInterval retry-interval]
[TransferDataSize {512|1024|2048|4096}]
[LifeofResource number-of-days]
[Compress {NO|COMPRESS|PACK|COMMON}]
[CommandDeletewithResource {YES|NO}]
[IDCommandDeleteTiming {RESOURCE_DELETE|MANUAL}]
[IDCommandDetailInf {YES|NO}]
[AutoAction {YES|NO}]
[EventType IM]
[DetailStatusInf {YES|NO}]
[OutPutFileOfDeliveryStatus file-name]
[OrderOfDeliveryStatus {M|R|G|S|A|K}]
[SystemLogOutputLevel {0|1|2|3}]
[MaxEntryofMessage maximum-message-count-for-message-log-file]
[ParallelTransfer{YES|NO}]
[InformLineDown{YES|NO}]
[FileTransferSleepTime sleep-time]
[FileTransferSleepInterval sleep-interval]
[ReceiveWaitTime file-transfer-response-timeout]
[EnableMultiServer {YES|NO}]
[MaxExecuteClients maximum-number-of-startup-clients]

(2) Details

ManagingHost host-name ~ <uppercase alphabetic characters, numerics, hyphens, underscores, or periods>((1 to 64 characters))
In host-name, specify the host name of the managing server or a relay system to which you want to connect the local UNIX machine. You must specify a host name specified in the /etc/hosts file. The value specified in this attribute becomes the default if the host file name is not specified in the packaging-information file or in the option of a package operation command (rdsft).
When ConnectionKind has been set to MASTER (connect to the relay system), specify the host name for the upper-level relay system.
If the managing server is running in a cluster environment, specify a host name that is compatible with the logic IP address defined by the cluster software. For details about the operations in a cluster environment, see the manual Job Management Partner 1/Software Distribution Manager Description and Administrator's Guide.

WorkstationType
In this attribute, specify MASTER, which indicates a relay system.

ProtocolKind TCP
Specify TCP as the communication protocol with the managing server.

ConnectionKind ~ {HOST|MASTER}<<HOST>>
Specify the connection destination of the relay system.

HOST
Connects the managing server.

MASTER
Connects a relay system.

ClientActionProtocol ~ {UDP|TCP}<<UDP>>
Specify the communication protocol with which to request an action from the subordinate relay system or client. If the relay system is in a DNS environment, the subordinate relay system or client must also be able to resolve the address of a host name with a domain name.

UDP
Uses the UDP protocol.

TCP
Uses the TCP protocol.
Specify TCP if the network has a firewall or router that rejects the UDP protocol.

ClientActionRetryCount retry-count ~ <unsigned integer>((1 to 9))<<3>>
In retry-count, specify the maximum number of times the system performs a retry if an action request to the subordinate relay system or client fails.
This attribute takes effect when TCP is set in the ClientActionProtocol attribute. If UDP is specified in the ClientActionProtocol attribute, the system ignores the ClientActionRetryCount attribute and performs a retry ten times within 240 seconds.

ClientActionRetryInterval retry-interval ~ <unsigned integer> ((10 to 1800))<<300>> (in seconds)
In retry-interval, specify the retry interval at which the system performs a retry if an action request to the subordinate relay system or client fails.
This attribute takes effect when TCP is set in the ClientActionProtocol attribute. If UDP is specified in the ClientActionProtocol attribute, the system ignores the ClientActionRetryCount attribute and performs a retry ten times within 240 seconds.

MaxConnectClients number-of-clients ~ <unsigned integer> ((1 to 9999)) <<30>>
In number-of-clients, specify the maximum number of distribution-destination systems that can concurrently be connected to a relay system. You can limit the number of connections to reduce the load on the relay system. The amount of memory required to connect clients depends on the number of clients to be connected (see 6.2.5 Memory requirement). Check the amount of memory available in the environment, and specify a proper value.
If the specified number of connections is exceeded, connection requests from the lower-level system will be rejected. When the number of connections drops below the specified limit, a startup message is sent to the lower-level system to prompt a connection retry.
The number of connections set here takes effect when the command shown below is processed.
  • rdsatrup
  • rdscm
  • rdsft -c
  • rdsiddefine -dl
  • rdsiddefine -s
  • rdsidexec
  • rdsrf -a
  • rdsrf -q
  • rdsusnd
Control is suppressed if an error occurs during control of the number of distribution-destination systems. Control is restored when the system recovers from the error.
However, the number of distribution-destination systems set here does not include the number of workstations that are in the process of retry connection. Set the number of distribution-destination systems including such workstations in MaxExecuteClients (maximum number of clients that can start). However, if many distribution-destination systems do not respond to retry, a significant amount of time will be required for all distribution-destination systems to start up.

WatchTimeofStatus monitoring-interval ~ <unsigned integer>((0 to 1280))<<60>> (in minutes)
Specify the monitoring interval in minutes. Monitoring is used to periodically check whether or not the rdssnd command that sends job execution results has been sent from the lower-level system.
If you execute the rdssnd command with any value other than 0 specified, the results of processes that were completely executed within the specified monitoring time are sent to the managing server. If execution results have not been reported, they will not be sent to the managing server. If the results cannot be sent to the upper-level server, the results will be reported in the next report.
When you specify 0, monitoring will be performed only once immediately after configuration. No monitoring will be conducted thereafter. Note that execution results that were not sent from the client workstation during monitoring will not be sent to the managing server until the next command is executed.

WatchTimeofOrders monitoring-interval ~ <unsigned integer> ((0 to 60)) <<5>> (in minutes)
To check from the relay system for the presence of a distribution job on the managing server, specify in minutes the time interval for monitoring the execution of the command to check for the presence of a distribution job (rdsma command).
If you execute the rdsma command when this attribute is set, the system monitors the execution of the command at the specified intervals.
If you specify 0 in this attribute, the system monitors the command only once.

RetryCount retry-count ~ <unsigned integer>((1 to 9))<<3>> (number of times)
In retry-count, specify the maximum number of times the system performs a retry if it cannot establish an association (session) or connection with the managing server. If you specify 9, the system will retry endlessly.
The relay system not connected to the managing server ignores this attribute.

RetryInterval retry-interval ~ <unsigned integer>((1 to 99))<<1>> (in units of 10 seconds)
In retry-interval, specify the interval at which the system performs a retry if it cannot establish an association (session) or connection with the managing server. Specify a value in units of 10 seconds. For example, a retry interval of 1 indicates 10 seconds.
The relay system not connected to the managing server ignores this attribute.

ServerRetryCount retry-count ~ <unsigned integer>((1 to 9))<<3>> (number of times)
In retry-count, specify the maximum number of times the system performs a retry if it cannot establish a connection to the file transfer server facility. If you specify 9, the system will retry endlessly.

ServerRetryInterval retry-interval ~ <unsigned integer>((1 to 99))<<5>> (in seconds)
In retry-interval, specify the interval at which the system performs a retry if it cannot establish a connection to the file transfer server facility. Specify a value in seconds.

DeliveryRetryCount retry-count ~ <unsigned integer>((0 to 9))<<0>> (number of times)
In retry-count, specify the maximum number of times the system performs a retry if the package distribution command or the resource collection command fails. If you specify 0, the system does not perform a retry.

DeliveryRetryInterval retry-interval ~ <unsigned integer>((1 to 99))<<3>> (in units of 10 seconds)
In retry-interval, specify the interval at which the system performs a retry if a line failure occurs while the system is executing the package distribution command or the resource collection command. Specify a value in units of 10 seconds. For example, a retry interval of 1 indicates 10 seconds. Note that if the DeliveryRetryCount attribute is set to 0, the system ignores the DeliveryRetryInterval attribute.

TransferDataSize ~ {512 | 1024 | 2048 | 4096}<<1024>>
Specify the size of the buffer for exchanging data with the managing server or with the upper-level system in a TCP/IP environment.
Select one of the buffer size, considering the network environment including the connected managing server or the upper-level system.

LifeofResource number-of-days ~ <unsigned integer>{1 to 999}<<1>> (in days)
In number-of-days, specify the number of days for storing resources. If the expiration date of a resource is not specified in the packaging-information file, the resource is stored in the temporary package storehouse for the number of days specified in this attribute. The default is one day (until the day after registration).

Compress ~ {NO | COMPRESS | PACK | COMMON}<<NO>>
Specify the compression method for the packages you want to register.

NO
Does not compress the packages.

COMPRESS
Uses the compress command to compress the packages.

PACK
Uses the pack command to compress the packages.

COMMON
Compresses shared packages.
Specifying PACK has no effect on a package consisting of three or fewer blocks.
If compression of a package fails, the package is registered but is not compressed. The value specified in this attribute becomes the default for the packages for which the compression method is not specified in the packaging-information file.

CommandDeletewithResource ~ {YES | NO} <<NO>>
Specify whether or not to delete the jobs left in the local or relay system when the packages associated with those jobs are deleted from the temporary package storehouse.

YES
Deletes the jobs when the packages are deleted from the temporary package storehouse.

NO
Does not delete the jobs even when the packages are deleted from the temporary package storehouse.

IDCommandDeleteTiming ~ {RESOURCE_DELETE|MANUAL} <<RESOURCE_DELETE>>
The jobs that specify ID groups are disabled when the packages associated with the jobs are deleted from the temporary package storehouse. This attribute specifies how to delete the disabled jobs.

RESOURCE_DELETE
Automatically deletes the disabled jobs when the associated packages are deleted from the temporary package storehouse.

MANUAL
Does not automatically delete the disabled jobs when the associated packages are deleted from the temporary package storehouse. The user deletes the jobs at any time.

IDCommandDetailInf ~ {YES | NO}<<NO>>
Specify whether or not to report the execution result of a job that specifies an ID group to the managing server for each client.

YES
Reports the execution results for each client to the managing server.
Note the following:
  • You can use the execution result (job) reported to the managing server only to check the distribution status. Deleting jobs which are being executed will not cause jobs in the distribution-destination system to be deleted.
  • If the managing server program is a previous version, jobs which resulted in an EXEC_ERROR (startup failure during job execution) will be displayed as ERROR (execution failed).

NO
Executed results are not reported to the managing server.

AutoAction ~ {YES | NO}<<NO>>
Specify whether to send a JP1 event to JP1/Base if an error occurs or the status is changed.

YES
Sends a JP1 event.

NO
Does not send a JP1 event.
For details on the automatic action and other operations using JP1 events, see 8. JP1 Event Processing Using JP1/IM - Manager.

EventType IM

Specify IM as the type of JP1 events to be issued.

DetailStatusInf ~ {YES | NO}<<NO>>
Specify whether or not to report the details of the job execution status to the managing server.

YES
Reports the details of the job status. On the managing server side, job status is reported as follows:
  • If a distribution job that installs a package during system operation is executed, the end status of the job changes to 0x80 when transfer of the package to the distribution-destination system is completed.
  • When multiple distribution jobs are executed simultaneously, the end of execution is reported for each job.

NO
Does not send a detailed report.

OutPutFileOfDeliveryStatus file-name ~ <alphanumeric characters, slants, or periods> ((up to 128 characters))
Specifies the absolute path name of the file name containing the distribution status name. The following files will be assumed if this value is omitted:
  • For systems other than HP-UX:/NETMRDSEX1/DLV/DLVF
  • For HP-UX:/opt/NETMDMWEX1/DLV/DLVF
You can use either a library function or a command to reference the distribution status file. The file can contain distribution status until the file system becomes full. The superuser is the owner of the distribution status file.

OrderOfDeliveryStatus ~ {M | R | G | S | A | K}
Specify acquiring the distribution status of a job.

M
Transfer package to relay system job

R
Recovery job

G
Collect files from distribution-destination system job (normal collection job)

S
Collect files from client to relay system job

A
Acquire collected files from relay system job

K
Deletion of the temporary collection storehouse
If you want to specify more than one keyword in this attribute, use the following format:
OrderOfDeliveryStatus MRGSAK
The system always acquires the distribution status for the Install package job, Get installed package status job and the rdsusnd command.

SystemLogOutputLevel ~ {0 | 1 | 2 | 3}<<3>>
Specify the level of the messages to be output to the syslog file.

0
Does not output messages.

1
Outputs error messages and system error messages (the messages whose ID ends with E or S).

2
Outputs error messages, system error messages and warning messages (the messages ends with E, S or W).

3
Outputs all the messages to be output to syslog. The following messages are output to syslog:
  • Messages that are output if an error occurs except for incorrect specification of operands in commands
  • Messages that are output for normal operation when JP1/Software Distribution runs in the background
The following table lists the messages that the client services output to syslog.
Message ID Facility Level
KDDMnnnn-E LOG_USER LOG_ERR
KDDMnnnn-S LOG_USER LOG_CRIT
KDDMnnnn-W LOG_USER LOG_WARNING
KDDMnnnn-I LOG_USER LOG_INFO

Legend:
nnnn: A four-digit number is displayed as the message code.

MaxEntryofMessage maximum-message-count-for-message-log-file ~ <unsigned integer>((300 to 99999))<<300>>
Specify the maximum number of messages that can be stored in a message log file. If you change this value to be smaller than before, the message log file is edited beginning from the latest message up to the new maximum specified. Delete unnecessary messages to retain the necessary ones. Editing of the message log file temporarily requires additional disk space for the value specified here. Before specifying this item, make sure that there is sufficient free space on the disk.

ParallelTransfer ~ {YES | NO}<<NO>>
Specify whether to upload and download at the same time between the local machine and managing server or relay system that is connected.

YES
Upload and download at the same time.

NO
Does not upload and download at the same time.

InformLineDown ~ {YES | NO}<<NO>>
Specify whether to notify the managing server of communication errors at package distribution. If YES is set and a communication error occurs during communication with the lower-level system, the distribution managing server is not notified of any communication error that occurs after recovery from the error.

YES
Notify the managing server.

NO
Do not notify the managing server.
Note that if you specify NO, any LINE-DOWN status already recorded in the managing server when you specify NO will not return to KEEP after recovery.

FileTransferSleepTime ~ <unsigned integer> ((0 to 4294967295)) <<0>> (unit: milliseconds)
Specify a sleep time (milliseconds) during file transfer.
This attribute takes effect only when a non-zero value is set in FileTransferSleepInterval.
If the distribution-destination system is running JP1/Software Distribution SubManager or JP1/Software Distribution Client 06-52 or an earlier version, set a sleep time of 240,000 milliseconds or less.

FileTransferSleepInterval ~ <unsigned integer> ((0 to 4294967295)) <<0>> (unit: packets)
Specify the amount of data (packets) to transfer before the sleep time is initiated.
Each time the specified number of packets have been transferred, file transfer pauses for the time set in FileTransferSleepTime.
This attribute takes effect only when a non-zero value is set in FileTransferSleepTime.

ReceiveWaitTime file-transfer-response-timeout ~ <unsigned integer>((0 to 2147483647))<<240>> (unit: seconds)
Specifies in seconds the time to monitor for a response from the server after a file is sent. If there is no response from the server within the specified file transfer time, the system assumes that communication failed. If you specify 0, response is not monitored. Make sure the value you specify here is not less than the sleep time between transmissions (FileTransferSleepTime) set in the manager services settings file; otherwise, a communication error will occur when there is a pause between transmissions.

EnableMultiServer ~ {YES|NO}<<NO>>
Specifies whether or not to accept jobs executed at multiple upper-level systems. A system configuration that has multiple upper-level servers are called multiservers.

YES
Specify this when operating on a multiserver.
According to this specification, job execution results are sent to the host that downloaded the job, instead of the host specified in ManagingHost.
However, if the following errors occur, the job may still remain in the ERROR status after recovery from the error. If this happens, recreate the job at the managing server and then execute it.
  • A communication error occurred during communication with the relay system (except for the relay system set at the ManagingHost).
  • File transfer from the managing server was interrupted.

NO
Specify this when not operating on a multiserver.
At this setting, job execution results, even those downloaded from other hosts, may be sent to the ManagingHost.

MaxExecuteClients maximum-number-of-action-clients-to-start ~ <Signless Integer>((1 to 9999))
Sets the number of distribution-destination systems that can be connected simultaneously. Unlike MaxConnectClients (the number of distribution-destination systems that can be connected simultaneously), this value includes the number of distribution-destination systems involved with connection retries. However, if there are many distribution-destination systems that do not respond to retries, much time will be required to start the all distribution-destination systems. Decide whether or not to set based on the system status.
If this setting is omitted, the settings value from MaxConnectClients that specifies the number of distribution-destination systems that can be connected will be applied. (This value does not include the number of distribution-destination systems involved in connection retries.) If a value greater than the set value in MaxConnectClients is specified, the maximum specified for MaxConnectClients will be applied.

[Contents][Back][Next]

[Trademarks]

All Rights Reserved. Copyright (C) 2009, Hitachi, Ltd.