set format

[set prc_process_count=maximum-concurrent-server-processes]
[set prc_recovery_resident=Y|N]
[set prc_take_over_svpath=Y|N]
[set term_watch_time=abnormal-termination-check-expiration-time]
[set max_socket_descriptors=maximum-number-of-file-descriptors-for-
                            sockets]
[set term_watch_count=limit-of-consecutive-abnormal-terminations]
[set prc_prf_trace=Y|N]
[set prc_coresave_path=core-file-storage-path]
[set prc_corecompress=Y|N]
[set watch_time=maximum-response-waiting-time]

command format

[prcsvpath path-name]

The process service definition defines the execution environment to manage server processes executed under an OpenTP1 system environment.

prc_process_count=maximum-concurrent-server-processes~<unsigned integer> ((1-10000)) <<64>>

Specify the maximum number of server processes to be started concurrently within the same machine. Note that the maximum number of processes that can actually be started up depends on the machine that starts up the OpenTP1 system. When specifying the prc_process_count operand, consider the installed memory size and the maximum number of processes that can be started up by the machine. For this, include all resident and non-resident server processes (user servers and system services) to be operated within an OpenTP1 system.

Compute the total of all servers to be started if using a multi-server system. The number of system services is the sum of the log services, journal related services (checkpoint dump, system journal, and other journal services), MCFs, and MQ systems.

prc_recovery_resident=Y|N~<<Y>>

Specify whether to make resident or non-resident the partial-recovery process when a UAP terminates abnormally. If the process is made non-resident, it is started when the UAP goes down. In case there is not enough memory when operating OpenTP1, either specify Y or omit this operand. If you specify Y or omit this operand, you will get a single resident process for partial recovery.

Y

The partial-recovery process is made resident.

N

The partial-recovery process is made non-resident.

prc_take_over_svpath=Y|N~<<N>>

Specify whether to inherit the user server and command path within the node at a rerun.

Y

The user server and command path are inherited at a rerun.

N

The user server and command path are not inherited at a rerun.

When Y is specified and the user server or command path is changed, the change is retained until OpenTP1 is started normally.

The user server and command path are inherited only within the same node. They are not inherited to a different system when a system switchover occurs.

If the inheritance fails, a warning message is output and the processing is continued without inheriting the user server and command path changed when the previous online session.

term_watch_time=abnormal-termination-check-expiration-time~<unsigned integer> ((0-32767)) <<10>> (Unit: minutes)

Specify a period of time to monitor the number of times OpenTP1 terminates abnormally. Specify 0 to have no time check.

If the OpenTP1 system is shut down three times consecutively within the specified time, the process service outputs the KFCA00715-E message and halts the system startup or restart. In this case, remove the cause of the shutdown, and then enter the dcsetup command to delete and re-register OpenTP1 to the operating system, or enter the dcreset command.

When MANUAL2 is specified for the mode_conf operand of the system environment definition, this operand is effective only for consecutive shutdown of the process service.

max_socket_descriptors=maximum-number-of-file-descriptors-for-sockets~<unsigned integer> ((32-2032))

Specify the maximum number of file descriptors to be used for sockets by the processes under control of OpenTP1^#.

The processes under control of OpenTP1^# exchange the process information with the system servers or user servers through the TCP/IP communication using sockets. Therefore, you must change the maximum number of file descriptors for sockets depending on the number of UAP processes that run concurrently and the number of other nodes to communicate with.

#: OpenTP1 processes other than the MCF services (MCF manager service, MCF communication service, and application startup service). For the MCF services, see the sections on the system service information definition and the system service common information definition.

Use the following formula for calculating the maximum number of file descriptors for sockets:

(Number of UAP processes in the local node^#1 + number of nodes that request a service of the name service^#2 + number of system service processes^#3)/0.8

: Rounded up to the nearest whole integer.

#1: The number of UAP processes in the local node is the sum of the following values:

• Number of UAP processes in the local OpenTP1
• Number of transactions to be started concurrently by the CUP (value of the parallel_count operand specified in the client service definition)

#2: Add this only when using TP1/Multi. This value is the sum of the following values:

• Number of UAP processes in other nodes that call the dc_adm_get~ function for the local node
• Number of dcmstart, dcmstop, and dcndls commands to be entered for the multi-node area and multi-node subarea that contains the local node as a building block

#3: The number of system service processes in the local OpenTP1.

If the value specified for this operand is too small, the connection cannot be set with other processes under control of OpenTP1. The process terminates abnormally after outputting the KFCA00307-E error message.

The order of priority of the specified values is 1 > 2.

1. Process service definition
2. System common definition

If this operand is omitted, the system assumes the value in the system common definition.

term_watch_count=limit-of-consecutive-abnormal-terminations~<unsigned integer>((1-3))<<3>>

When you specify 1 to 3 in this operand, you can use the facility of specifying the limit of consecutive abnormal terminations for the OpenTP1 system.

The term_watch_time operand becomes valid when 3 is specified in this operand or this operand is omitted. Table 3-2 shows the relationship between this operand and the term_watch_time operand.

Table 3-2 Relationship between term_watch_count and term_watch_time operands

If the OpenTP1 system goes down the number of times specified in this operand, the process service outputs the KFCA00715-E message and cancels the system startup or restart. In this case, correct the error that caused the system failure, then use the dcsetup command to delete the registration in the operating system and re-register OpenTP1 or enter the dcreset command.

If the mode_conf operand in the system environment definition is MANUAL2, the specification of this operand is invalid except in the case where there are consecutive failures of the process service.

prc_prf_trace=Y|N~<<Y>>

Specify whether to acquire the trace information for performance verification when a process is generated or disappears.

Specify whether to acquire the event trace for the process service.

Y

The event trace for the process service is acquired.

N

The event trace for the process service is not acquired.

When Y is specified for this operand, the event trace for verifying the process service operation is acquired. For details about event IDs, see the manual OpenTP1 Operation.

The acquired event trace information is stored in the following files:

• 0xb001 to 0xb003: Trace files in $DCDIR/spool/dcprfinf
• 0xb010 and following events: Trace files in $DCDIR/spool/dcprcinf

For the event trace, the information about process generation and disappearance is acquired at the following times:

1. By a child process when a process is generated (the event ID is 0xb001)
2. When a process disappears (the event ID is 0xb002)
3. By the parent process when a process is generated (the event ID is 0xb003)

When you specify the -d option when executing the prfed command, as the data to be displayed, method 1 acquires the server name of the generated process, method 2 acquires the ID of the process that disappeared and the end status, and method 3 acquires the generated process ID. To output the acquired trace information using the prfed command, specify the -d option. When you specify the -d option, trace information is output as follows:

For method 1:

Offset +0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +a +b +c +d +e +f ASCII_code +0000 aa aa aa aa aa aa aa aa AAAAAAAA

• aa....aa: Server name of the generated process (hexadecimal number)
• AA....AA: Server name of the generated process (ASCII code)

For method 2:

Offset +0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +a +b +c +d +e +f ASCII_code +0000 bb bb bb bb cc cc cc cc BBBBCCCC

• bb....bb: ID of the process that disappeared (hexadecimal number)
• BB....BB: ID of the process that disappeared (ASCII code)
• cc....cc: End status of the process that disappeared (hexadecimal number)
• CC....CC: End status of the process that disappeared (ASCII code)

For method 3:

Offset +0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +a +b +c +d +e +f ASCII_code +0000 dd dd dd dd DDDD

• dd....dd: Generated process ID (hexadecimal number)
• DD....DD: Generated process ID (ASCII code)

For details about the prfed command, see the manual OpenTP1 Operation.

When you specify N for the prf_trace operand in the system common definition, the prc_prf_trace operand is invalid.

prc_coresave_path=core-file-storage-path~<path name> <<$DCDIR/spool/save>>

Specify the absolute path name for the directory that stores the core file of an OpenTP1 process.

For a user server, the result of editing the UAP trace is also stored in the directory specified using this operand.

Up to 63 characters can be specified for a path name.

If the path name specified using this operand contains an error or if you do not specify this operand, $DCDIR/spool/save is assumed as the storage destination of the core file.

The OpenTP1 system administrator must have the permission to write the core file in the directory specified in this operand. Otherwise, the core file may not be written in the specified directory.

The core file of the process server may be output to $DCDIR/spool/save regardless of the specification of this operand.

If you change the value specified for this operand, you need to execute the dcreset command.

prc_corecompress=Y|N~<<N>>

Specify whether to allow OpenTP1 to automatically compress the core file of the OpenTP1 process when storing the core file.

Y

OpenTP1 automatically compresses the core file of the OpenTP1 process when storing the core file.

N

OpenTP1 does not automatically compress the core file of the OpenTP1 process when storing the core file.

The core file is not compressed when the operand is not specified.

Compression processing is performed using a standard command provided by the operating system if a user application program goes down while OpenTP1 is online. Therefore, when you specify Y for this operand and a user application program goes down, the entire performance may degrade depending on the performance of the standard compression command provided by the operating system.

When you specify Y for this operand, a file containing no data with a file name before compression is created in the directory specified for storing the core file. Do not delete this file since it may be used to manage the generations of information files that are backed up.

When you change the value specified for this operand, you need to execute the dcreset command.

watch_time=maximum-response-waiting-time~<unsigned integer> ((0-65535)) (Unit: seconds)

Specify the maximum waiting time between sending a service request and receipt of a response when communicating between processes by RPC.

OpenTP1 may suspend termination processing for the length of time specified in this operand. Therefore, if you specify a large value, the termination processing of OpenTP1 may take some time.

If no response is received within the specified time, RPC returns a transmission timeout error.

Specify 0 if the system is to wait for a response. When you specify 0, OpenTP1 may not terminate.

If this operand is omitted, the value specified with the watch_time operand of the system common definition is assumed.

Use the default of the watch_time operand of the system common definition for this operand.

We recommend that you do not change the operand unless special tuning is necessary.

If a value that is mush greater or smaller than the default of the watch_time operand of the system common definition is specified, a failure may occur causing OpenTP1 to go down.

Described on the following page.