Operands

dcnamuse=Y|N ~<<N>>
Specifies whether or not RPCs that use the name service are used.

Y: Use RPCs that use the name service.

N: Do not use RPCs that use the name service.

The dcnamuse, dcrapdirect, and dcscddirect operands are mutually exclusive. You cannot specify Y for more than one of these operands at the same time; doing so results in a definition error.
The TP1/Web connection facility can be used only if you specify Y in the dcrapdirect operand. Defining the dcweburl operand and specifying dcscddirect=Y and dcrapdirect=Y results in a definition error.
dcnamport=port-number-of-name-service
~<unsigned integer> ((5001~65535)) <<10000>>
Specifies the port number of the name service that was specified in the name_port operand of the system common definition for the OpenTP1 to which the client belongs. This value takes effect on a host for which no port number is specified in the dchost operand.
dchost=host-name-of-TP1/Server-as-a-gateway ~<character string>
Specifies the host name of the TP1/Server as a gateway and the port number of the RAP-processing listener, schedule service, or name service. If you use a delimiter (,), you can specify multiple TP1/Servers. To issue RPCs via a RAP-processing listener, specify the host name of the RAP-processing listener or the host name of the firewall.
If both the dchost operand and the dcweburl operand are specified, the dcweburl operand takes precedence.

Format

When dcrapdirect=Y is specified:

host-name[:port-number-of-RAP-processing-listener][,host-name[:port-number-of-RAP-processing-listener],...]

When dcscddirect=Y is specified:

host-name[:port-number-of-schedule-service][,host-name[:port-number-of-schedule-service],...]

When dcnamuse=Y is specified:

host-name[:port-number-of-name-service][,host-name[:port-number-of-name-service],...]

- host-name ~<character string>

- port-number ~<unsigned integer> ((5001~65535))

You can specify a maximum of 256 characters. Do not enter any blank characters (spaces or tabs), except after a delimiter (,). For the host name, you can also specify the IP address expressed in decimal dot format.

If you specify more than one host name for TP1/Server and a TP1/Server error is detected, TP1/Client/J attempts swapping by referencing the next TP1/Server specified in the dchost operand. However, if RPCs are issued via a RAP-processing listener, TP1/Client/J attempts swapping by referencing the next TP1/Server specified in the dchost operand only when dcrapautoconnect=Y is also specified in the TP1/Client/J environment definition or DCRPC_RAP_AUTOCONNECT is specified in the setDccltextend method.

For the port number, if you have specified dcrapdirect=Y, specify the port number of the RAP-processing listener. If you have specified dcscddirect=Y, specify the port number of the schedule service. If you have specified dcnamuse=Y, specify the port number of the name service.

If the port number is omitted and dcrapdirect=Y is specified, TP1/Client/J assumes the value of the dcrapport operand; if dcscddirect=Y is specified, TP1/Client/J assumes the value of the dcscdport operand; if dcnamuse=Y is specified, TP1/Client/J assumes the value of the dcnamport operand.
dcwatchtim=maximum-time-to-wait-for-a-response
~<unsigned integer> ((0~65535)) <<180>> (seconds)
Specifies the maximum time to wait for a response from the service since a service request was issued from the CUP to the TP1/Server.
If no response is returned within the specified time, control is returned to the CUP with an error.
If 0 is specified, TP1/Client/J waits indefinitely for a response.
The specification of the dcwatchtim operand is applied to the following APIs (methods):
- closeConnection method
- openConnection method
- rpcCall method (except for non-response type RPCs that use the scheduler direct facility)
- rpcCallTo method (for synchronous-response RPCs)
- trnBegin method
- trnChainedCommit method
- trnChainedRollback method
- trnUnchainedCommit method
- trnUnchainedRollback method
dccltinquiretime=maximum-permanent-connection-inquiry-interval
~<unsigned integer> ((0~1048575)) (seconds)
Specifies the maximum amount of time from one inquiry to another that is issued from a CUP to the server. The maximum permanent connection inquiry interval is the timer value that is monitored by the RAP-processing server. If there is no inquiry within the specified time, the RAP-processing server releases the permanent connection forcibly. If the maximum permanent connection inquiry interval is reached within a transaction, TP1/Client/J rolls back the corresponding transaction forcibly.
If you specify 0 in this operand, TP1/Client/J waits indefinitely for a response from the CUP. If you omit this operand, the rap_inquire_time operand in the RAP-processing listener service definition takes effect.
dcwatchtiminherit=Y|N ~<<N>>
Specifies whether or not the RAP-processing server is to inherit the CUP's maximum amount of time to wait for a response when an RPC that uses the remote API facility is issued.

Y: The RAP-processing server is to inherit the CUP's maximum time to wait for a response.

N: The RAP-processing server is not to inherit the CUP's maximum time to wait for a response.
dcwatchtimrpcinherit=Y|N ~<<N>>
Specifies whether or not the TP1/Server is to inherit the CUP's maximum amount of time to wait for a response.

Y: TP1/Server is to inherit the CUP's maximum time to wait for a response.

N: TP1/Server is not to inherit the CUP's maximum time to wait for a response.
dccltdelay=maximum-communication-delay-time
~<unsigned integer> ((0~65535)) <<0>> (seconds)
Taking into account the overhead involved with communication between the CUP and the RAP-processing server, you can specify this operand to end response monitoring in the RAP-processing server system earlier than in the TP1/Client/J system. This definition ends response monitoring in the RAP-processing server system early by the specified amount of time in order to avoid missing a message that may arrive after a timeout occurs in the TP1/Client/J system.
This definition is applicable only when dcrapdirect=Y and dcwatchtiminherit=Y are specified in the TP1/Client/J environment definition.
TP1/Client/J ignores the dccltdelay operand when dcwatchtim=0 is specified in the TP1/Client/J environment definition, 0 is specified in the setDcwatchtim method, or dcwatchtiminherit=N is specified in the TP1/Client/J environment definition. If the dcwatchtim operand value minus the dccltdelay operand value equals 0 or a negative value, TP1/Client/J also ignores the dccltdelay operand, in which case TP1/Client/J assumes a value of 1.
dcselint=reply-text-monitoring-interval ~<unsigned integer> ((1~65535)) <<100>> (milliseconds)
This operand is provided only to provide compatibility with the source of the previous version. Currently, TP1/Client/J ignores this operand, if specified.
dccltextend=00000000|00000001
Specifies the facility extension level of TP1/Client/J.

00000000: Do not extend the TP1/Client/J facility.

00000001: Send the IP address of the local CUP to the service when the rpcCall method is called. You specify this value if you need to obtain the address of the CUP by having the called service execute the dc_rpc_get_callers_address() function.
dccache=maximum-entries-count-in-name-cache ~<unsigned integer> ((2-256)) <<8>>
Specifies the maximum number of entries in the cache that is used to cache service information used by TP1/Client/J. This cache is managed by the LRU method. When the number of entries in cache exceeds the specified value, TP1/Client/J deletes the oldest service information that was called in the past.
dchostselect=Y|N ~<<N>>
Specifies whether or not to randomly select a TP1/Server as a gateway. This definition is applicable only when more than one TP1/Server is specified as a gateway in the dchost operand.
Y: Select a TP1/Server as a gateway at random.
N: Do not select a TP1/Server as a gateway at random.

When Y is specified

If an error is detected while an inquiry is being made to the name service of the TP1/Server as a gateway, TP1/Client/J randomly selects another TP1/Server for swapping.

When N is specified

If an error is detected while an inquiry is being made to the name service of the TP1/Server as a gateway, TP1/Client/J attempts to switch to the next TP1/Server specified in the dchost operand.

The erroneous TP1/Server becomes a candidate for a retry-target host for the next RPC.
If dcscddirect=Y is specified in the TP1/Client/J environment definition, swapping of gateway TP1/Servers occurs when transmission to the corresponding port number fails.
dcscddirect=Y|N ~<<N>>
Specifies whether or not RPCs that use the scheduler direct facility without querying information to the name service of TP1/Server are used.

Y: Use RPCs that use the scheduler direct facility.

N: Do not use RPCs that use the scheduler direct facility.

If you have specified the port number of the schedule service in the dcscdport operand of the TP1/Client/J environment definition, TP1/Client/J uses that port number for querying. The dcrapdirect, dcscddirect, and dcnamuse operands are mutually exclusive. You cannot specify Y for more than one of these operands at the same time; doing so results in a definition error.
The TP1/Web connection facility can be used only if you specify Y in the dcrapdirect operand. Defining the dcweburl operand and specifying dcscddirect=Y and dcrapdirect=Y results in a definition error.
dcscdport=port-number-of-schedule-service
~<unsigned integer> ((5001~65535))
Specifies the port number of the schedule service that has been specified in the scd_port operand of the schedule service definition for the OpenTP1 to which the client belongs. This value takes effect on a host for which no port number has been specified with the dchost operand. This operand is applicable only when dcscddirect=Y is specified in the TP1/Client/J environment definition. For details, see the dcscddirect operand of the TP1/Client/J environment definition.
dcscdloadpriority>=Y|N ~<<N>>
Specifies whether or not the TP1/Server that accepts service requests as a gateway has priority for load balancing.
Y: The TP1/Server that accepts service requests as a gateway has priority for load balancing.
N: The TP1/Server that accepts service requests as a gateway does not have priority for load balancing. TP1/Server balances the workload between nodes.
This definition is applicable only when RPCs that use the scheduler direct facility are issued (dcscddirect=Y is specified in the TP1/Client/J environment definition).
dcrapdirect=Y|N ~<<Y>>
Specifies whether or not the facility for directly querying the RAP-processing listener without querying the information to the TP1/Server's name service (remote API facility) is used.
To put this operand into effect, specify the port number of the RAP-processing listener in the dcrapport operand of the TP1/Client/J environment definition.
The dcrapdirect, dcscddirect, and dcnamuse operands are mutually exclusive. You cannot specify Y for more than one of these operands at the same time; doing so results in a definition error.
dcrapport=port-number-of-RAP-processing-listener ~<unsigned integer> ((5001-65535))
Specifies the port number of the RAP-processing listener of the TP1/Server to which the client belongs or the port number of the firewall. This operand value takes effect on a host for which no port number has been specified with the dchost operand.
This operand is applicable only when dcrapdirect=Y is specified in the TP1/Client/J environment definition. For details, see the dcrapdirect operand of the TP1/Client/J environment definition.

dcrapautoconnect=Y|N ~<<N>>
Specifies whether or not a permanent connection is established automatically between a CUP and a RAP-processing server.

Y: Establish a permanent connection automatically the next time the method is executed. The target to which a permanent connection establishment request is issued is the RAP-processing listener that is specified in the dchost operand of the TP1/Client/J environment definition.

N: Do not establish a permanent connection automatically.

Operation of various methods based on the connection establishment mode

Table 5-2 shows connection establishment processing of the non auto connect mode and the auto connect mode.

Table 5-2 Operation of various methods based on the connection establishment mode

Value specified in the dcrapautoconnect operand	openConnection() method is issued	openConnection(host, port) method is issue	openConnection(url, flags) method is issued	rpcCall() is issued
N	If the dchost operand is specified, a connection is established to the RAP-processing server specified by the dchost operand. If the dcweburl operand is specified, a connection is established to the URL specified by the dcweburl operand. If neither the dchost nor dcweburl operand is specified, ErrHostUndefException is returned.	A connection is established to the RAP-processing server specified by the host and port parameters.	A connection is established to the URL specified by the url parameter.	ErrProtoException is returned.
Y	ErrProtoException is returned.	ErrProtoException is returned.	ErrProtoException is returned.	If the dchost operand is specified, a connection is established to the RAP-processing server specified by the dchost operand. If the dcweburl operand is specified, a connection is established to the URL specified by the dcweburl operand. If neither the dchost nor dcweburl operand is specified, ErrHostUndefException is returned.

Notes

The value of the dcrapautoconnect operand can be dynamically changed by the setRpcextend method.
The definition is read by the rpcOpen method.
The definition information read by the rpcOpen method is discarded by the rpcClose method.
When an openConnection method that specifies a URL as an argument is executed, the specified URL applies only to the openConnection method that specified the URL. Once this method has executed, the URL specified in this parameter is not used for another openConnection method (which has no parameter specified) that executes after the closeConnection method has been used to close the connection.

dcerrtrace=Y|N ~<<N>>
Specifies whether or not the CUP collects an error trace.

Y: Collect error trace.

N: Do not collect error trace.

If you are using the CUP as a Java applet, do not specify Y; if Y is specified, operations cannot be guaranteed.
dcerrtracepath=directory-for-error-trace-file ~<character string>
Specifies the path name of the directory in which error trace files are to be created.
This operand is applicable only when dcerrtrace=Y is specified in the TP1/Client/J environment definition.
dcerrtracesize=error-trace-file-size ~<unsigned integer> ((4096-1048576)) <<4096>> (bytes)
Specifies the size of an error trace file.
dcmethodtrace=Y|N ~<<N>>
Specifies whether or not the CUP collects a method trace.

Y: Collect method trace.

N: Do not collect method trace.

If you are using the CUP as a Java applet, do not specify Y; if Y is specified, operations cannot be guaranteed.
dcmethodtracepath=directory-for-method-trace-file ~<character string>
Specifies the path name of the directory in which method trace files are to be created.
This operand is applicable only when dcmethodtrace=Y is specified in the TP1/Client/J environment definition.
dcmethodtracesize=method-trace-file-size ~<unsigned integer> ((4096-1048576)) <<4096>> (bytes)
Specifies the size of a method trace file.
dcuaptrace=Y|N ~<<N>>
Specifies whether or not the CUP collects a UAP trace.

Y: Collect UAP trace.

N: Do not collect UAP trace.

If you are using the CUP as a Java applet, do not specify Y; if Y is specified, operations cannot be guaranteed.
dcuaptracepath=directory-for-UAP-trace-file ~<character string>
Specifies the path name of the directory in which UAP trace files are to be created.
This operand is applicable only when dcuaptrace=Y is specified in the TP1/Client/J environment definition.
dcuaptracesize=UAP-trace-file-size ~<unsigned integer> ((4096-1048576)) <<4096>> (bytes)
Specifies the size of a UAP trace file.
dcdatatrace=Y|N ~<<N>>
Specifies whether or not the CUP collects a data trace.

Y: Collect data trace.

N: Do not collect data trace.

If you are using the CUP as a Java applet, do not specify Y; if Y is specified, operations cannot be guaranteed.
dcdatatracepath=directory-for-data-trace-file ~<character string>
Specifies the path name of the directory in which data trace files are to be created.
This operand is applicable only when dcdatatrace=Y is specified in the TP1/Client/J environment definition.
dcdatatracesize=data-trace-file-size ~<unsigned integer> ((4096-1048576)) <<4096>> (bytes)
Specifies the size of a data trace file.
dcdatatracemaxsize=maximum-data-size-for-data-trace ~<unsigned integer> ((16-1048576)) <<128>> (bytes)
Specifies the maximum size of data per data trace item.
dcclttrstatisitem=statistical-information-item[,statistical-information-item]...
Specifies one or more of the character strings listed below to indicate the types of statistical information to be acquired for transaction branches. This operand is applicable only when a transaction is started from the CUP.

nothing

Do not acquire statistical information.

base

Acquire the following information as basic information:

- Transaction branch identifier

- Transaction branch determination result

- Transaction branch execution process type

- Name of the server executing a transaction branch

- Name of the service executing a transaction branch

executiontime

Acquire the basic information and the transaction branch execution time.

cputime

Acquire the basic information and the transaction branch CPU time.

If you specify nothing, you cannot specify any other value. If you specify nothing together with another value, nothing will be ignored.
To acquire statistical information about transactions, specify one of the following settings:
- trn_tran_statistics=Y in the transaction service definition
- -s option in the trnstics command
If you omit this definition, the trn_statistics_item operand specification in the RAP-processing listener service definition takes effect.
dcclttroptiitem=transaction-optimization-item[,transaction-optimization-item]...
Specifies one or more of the character strings listed below to indicate the optimization items to be used to improve the performance of a global transaction that consists of multiple user servers. This operand is applicable only when a transaction is started from the CUP.

base

Optimize the entire synchronization point acquisition processing (prepare, commit, and rollback processing). Because OpenTP1 uses the two-phase commit method for transaction control, four process-to-process communications are required for commit control between two transaction branches.

If all of the following conditions are satisfied, the four process-to-process communications required for commit control are eliminated by having a parent transaction branch execute commit processing for its child transaction branches:

- Both parent and child transaction branches are under the same OpenTP1.

- The parent transaction branch is using a synchronous-response RPC to call its child transaction branches.

- The object for the XA interface of the resource manager that is accessed by a child transaction branch is also linked to its parent transaction branch.

asyncprepare

If optimization of the entire synchronization point acquisition processing is not possible because the conditions for base are not satisfied, optimize the prepare processing.

If all of the following conditions are satisfied, two of the process-to-process communications are eliminated by executing preparatory processing before an RPC is returned when a child transaction branch issues a service request using the RPC issued by its parent transaction branch:

- Optimization with base specified is not available.

- The parent transaction branch is using a synchronous-response RPC to call its child transaction branch.

Note that this optimization increases the response time of a synchronous-response RPC that is issued by the parent transaction branch. For the child transaction branch, the interval from prepare to commit processing increases (because the transaction cannot be determined without an instruction from the parent transaction branch). Therefore, if the OpenTP1 system of the parent transaction branch is shut down and communications are interrupted between the transaction branches, swapping of journal files and validation of the checkpoint dump file are delayed and the OpenTP1 system of the child transaction branch may also be shut down.

You can specify more than one transaction optimization item, in which case the following priority rule applies:
```
1 > 2
```
1: base
2: asyncprepare
If you omit this definition, the trn_optimum_item operand specification in the RAP-processing listener service definition takes effect.
dcclttrwatchtime=maximum-communication-wait-time-during-transaction-synchronization-point-processing
~<unsigned integer> ((1~65535)) (seconds)
Specifies the maximum wait time for communications between transaction branches (prepare, commit, or rollback instruction or response) during transaction synchronization point processing. This operand is applicable only when a transaction is started from the CUP.
If no instruction or response is received within the specified time and the corresponding transaction branch is the first phase of a two-phase commit, TP1/Client/J rolls back the transaction branch; if the first phase has already been completed, TP1/Client/J uses a system process of the transaction service to retry transaction determination processing.
When this definition is omitted, the trn_watch_time operand in the RAP-processing listener service definition takes effect.
dcclttrrbinfo=no|self|remote|all
Specifies whether or not information about the cause of the rollback is logged when a transaction branch rolls back. This operand is applicable only when a transaction is started from the CUP.

no

Do not acquire rollback information.

self

Log rollback information only for those transaction branches that caused rollback.

remote

Log rollback information not only for those transaction branches applicable to self but also for those that received rollback requests from transaction branches in remote nodes.

all

Log rollback information not only for those transaction branches applicable to remote but also for those that received rollback requests from transaction branches in the local node.

When this definition is omitted, the trn_rollback_information_put in the RAP-processing listener service definition takes effect.
dcclttrlimittime=maximum-executable-time-for-transaction-branch
~<unsigned integer> ((0~65535)) (seconds)
Specifies the maximum executable time for a transaction branch. This operand is applicable only when a transaction is started from the CUP.
TP1/Client/J sets automatically the timeout values for the rpcCall method and synchronization point processing as follows, so that the amount of time from the start of a transaction branch to the end of synchronization point processing does not exceed the time specified in this operand:
- Timeout value for the rpcCall method
  If K >= value of this operand: Timeout occurs and TP1/Client/J returns control with an error without executing the request processing.
  If K < value of this operand and (value of this operand - K) >= W: TP1/Client/J uses W as the timeout value.
  If K < value of this operand and (value of this operand - K) < W: TP1/Client/J uses (value of this operand - K) as the timeout value.
  K and W indicate the following:
  K: Current time - transaction branch start time
  W: dcwatchtim operand value
  >=: Equal to or greater than
- Timeout value for communications within synchronization point processing
  If K >= value of this operand: TP1/Client/J sets the timeout value to 1 second.
  If K < value of this operand and (value of this operand - K) >= W: TP1/Client/J uses W as the timeout value.
  If K < value of this operand and (value of this operand - K) < W: TP1/Client/J uses (value of this operand - K) as the timeout value.
  K and W indicate the following:
  K: Current time - transaction branch start time
  W: dcclttrwatchtime operand value (if the dcclttrwatchtime operand is omitted, the dcwatchtim operand value takes effect)
  >=: Equal to or greater than
If time is used for processing other than waiting for data reception, the transaction branch may not be terminated within the specified operand value.
If the time specified in this operand elapses before synchronization point processing begins, the corresponding transaction is rolled back.
When 0 is specified, TP1/Client/J does not monitor the time.
When this definition is omitted, the trn_limit_time operand in the RAP-processing listener service definition takes effect.
dcclttrrbrcv=Y|N
Specifies whether or not a rollback completion report is received after a rollback instruction is sent to the RPC destination transaction branch. This operand is applicable only when a transaction is started from the CUP.
Y: Receive a rollback completion report.
N: Do not receive a rollback completion report.
When N is specified, TP1/Client/J terminates the local transaction branch without receiving a rollback completion report from the RPC destination transaction branch (without waiting for completion of rollback processing on the RPC destination transaction branch).
When this definition is omitted, the trn_rollback_response_receive operand in the RAP-processing listener service definition takes effect.
dcclttrrecoverytype=type1|type2|type3
Specifies the synchronization point processing method to use in the event of a UAP error. This operand is applicable only when a transaction is started from the CUP.
If an RPC results in a timeout and the address of the target process is unresolved or the UAP executing a transaction is shut down, transaction determination may take time because smooth communications cannot be achieved between transaction branches.
For this operand, select one of the three transaction synchronization point processing methods listed below as appropriate to one of the following types of errors that may occur:

(Error 1) When an RPC results in a timeout:

In this case, the source transaction branch that issued the RPC cannot send a transaction synchronization point message to the RPC destination transaction branch because it does not know which process is executing the service request. As a result, both source and target transaction branches are placed in transaction synchronization point message wait status, requiring time for transaction determination.

(Error 2) When the RAP-processing server shuts down before receiving a response from the RPC:

In this case, the source transaction branch that issued the RPC cannot send a transaction synchronization point message to the RPC destination transaction branch because it does not know which process is executing the service request. As a result, the RPC destination transaction branch is placed in transaction synchronization point message wait status, requiring time for transaction determination.

(Error 3) When the RAP-processing server and the UAP to which the RPC was issued shut down at about the same time after a response from the UAP to which the RPC was issued is received:

In this case, the transaction recovery process that inherited each transaction branch sends a transaction synchronization point message to a nonexistent UAP because it does not know that the remote UAP process has been shut down, requiring time for transaction determination.

The following describes the processing method indicated by each specification value:

type1

If (Error 1) occurs, both the source transaction branch that issued the RPC and the RPC destination transaction branch determine transactions when the process for receiving a transaction synchronization point message time out.

If (Error 2) occurs, the source transaction branch that issued the RPC determines the transaction without sending a transaction synchronization point message to the RPC destination transaction branch. The RPC destination transaction branch determines the transaction when the process for receiving a transaction synchronization point message times out.

If (Error 3) occurs, both the source transaction branch that issued the RPC and the RPC destination transaction branch determine transactions when the process for receiving a transaction synchronization point message times out.

type2

Same as type1 if (Error 1) occurs and a transaction is committed.

If (Error 1) occurs and a transaction is rolled back or if (Error 2) occurs, the source transaction branch that issued the RPC sends a transaction synchronization point message to the transaction service process on the node where the RPC destination transaction branch exists, and then determines the transaction. The transaction service process that has received the transaction synchronization point message sends a transaction synchronization point instruction to the process that is processing the applicable transaction branch.

If (Error 3) occurs, both the source transaction branch that issued the RPC and the RPC destination transaction branch determine transactions when the process for receiving a transaction synchronization point message times out.

type3

Same as type1 if (Error 1) occurs and a transaction is committed.

If (Error 1) occurs and a transaction is rolled back or if (Error 2) or (Error 3) occurs, the source transaction branch that issued the RPC sends a transaction synchronization point message to the transaction service process on the node where the remote transaction branch exists. The transaction service process that has received the transaction synchronization point message sends a transaction synchronization point instruction to the process that is processing the applicable transaction branch.

In the following cases, even if you specify type2 or type3 in this operand, transaction determination processing may take time:
- The status of the UAP to which the RPC was issued changed during RPC execution (such as an increase in the workload, UAP termination, or UAP shutdown) and a service request was forwarded to the same UAP on another node.
- The version of the remote OpenTP1 server does not support this option.
- The remote transaction branch is taking time for processing other than the transaction synchronization point message reception processing.
When this definition is omitted, the trn_partial_recovery_type in the RAP-processing listener service definition takes effect.
dcclttrexptm=expiry-time-in-transaction-branch
~<unsigned integer> ((0~65535)) (seconds)
Specifies the maximum transaction branch processing time. This operand is applicable only when a transaction is started from the CUP.
If the transaction branch is not completed within the specified time, TP1/Client/J terminates abnormally the process of that transaction branch and rolls it back. If 0 is specified, TP1/Client/J does not monitor the time.
When this operand is omitted, the trn_expiration_time operand in the RAP-processing listener service definition takes effect.
To specify whether or not this value includes the processing time of transaction branches that are executed by other processes when the RPC facility is used, use the dcclttrexpsp operand in the TP1/Client/J environment definition.
dcclttrcputm=transaction-branch-CUP-monitoring-time
~<unsigned integer> ((0~65535)) (seconds)
Specifies the CPU time allowed for a transaction branch before synchronization point processing starts. This operand is applicable only when a transaction is started from the CUP.
If 0 is specified, TP1/Client/J does not monitor the CUP time.
If the specified time is exceeded, TP1/Client/J terminates abnormally the process of the corresponding transaction branch and rolls it back.
When this operand is omitted, the trn_cpu_time operand in the RAP-processing listener service definition takes effect.
dcclttrexpsp=Y|N|F
Specifies whether the monitoring time includes the time of the following types of processing when transaction branch processing is monitored:
1. Time required for a transaction branch being monitored to call another transaction branch using the RPC facility and wait for completion of that processing
2. Time required for a server UAP called by a chained RPC to wait for the next service request
3. Time required for a transaction branch being monitored to execute result reception processing after using a non-response type RPC to call another transaction branch
The following describes each specification value:

Y: Include the monitoring time for 1, 2, and 3.

N: Include the monitoring time for 3 only.

F: Do not include the monitoring time for 1, 2, or 3.

When this operand is omitted, the trn_expiration_time_suspend operand in the RAP-processing listener service definition takes effect.
dcsndrcvtype=DCCLT_ONEWAY_SND|DCCLT_ONEWAY_RCV|DCCLT_SNDRCV
Specifies the environment to initialize when the TCP/IP communication facility is used.
These operands must be specified for you to use the TCP/IP communication facility.

DCCLT_ONEWAY_SND: Environment for unidirectional sending of messages

DCCLT_ONEWAY_RCV: Environment for unidirectional receiving of messages

DCCLT_SNDRCV: Environment for bidirectional sending and receiving of messages
dcrcvport=port-number-of-the-receiving-CUP
~<unsigned integer>((1~65535)) <<11000>>
Specifies the port number of the CUP that is receiving the message sent from the MHP.
Specify the same value as the port number used by the MHP for sending messages. If you are executing several processes and threads concurrently on the same machine, specify different port numbers for each process or thread.
Do not specify a port number that is being used by the OS or by any other program even if it is available. If such a port number is specified, correct reply data may not be received. Note that different OSs use different port numbers; for details, see the manual for the applicable OS manual.
dcsndhost=host-name-of-node-on-which-MHP-to-be-connected-exists~<character string>
Specifies the host name of the node on which the connection-target MHP exists when a connection is being established for a message to be sent from a CUP.
You can also specify the host name in the form of a decimal-plus-dot IP address.
dcsndport=port-number-of-MHP-to-be-connected
~<unsigned integer> ((1~65535)) <<12000>>
Specifies the port number of the connection-target MHP when a connection is being established for a message to be sent from a CUP. Specify a port number different from the one specified in the dcnamport operand.
dcsockopenatrcv=Y|N ~<<N>>
Specifies the trigger for opening a reception socket (the trigger to start waiting to receive a connection from a remote computer) when sending and receiving is performed over a single connection while the TCP/IP communication facility is being used.
If you do not specify this operand, the receive socket is opened when the rpcOpen method is executed.
Y: Opens a receive socket if a connection was not established when the cltReceive method was executed.
N: Opens a receive socket when the rpcOpen method is executed.
dcweburl=URL-of-TP1/Web-whose-service-is-requested ~<path-name>
Specifies the URL of the TP1/Web whose service is being requested.
Specify the protocol, Web server, the CGI name of the prompter, the TP1/Web service name, and other information in the form of a URL.
If both the dchost operand and the dcweburl operand are specified, the dcweburl operand takes precedence.
Note, also, that the dcscddirect and dcnamuse operands cannot be used if the TP1/Web connection facility is used. If the dcweburl operand is defined, and dcscddirect=Y and dcnamuse=Y are specified, a definition error results.
dccltdbgtrcfilecount=maximum-number-of-debug-trace-files
~<unsigned integer> ((0~256)) <<0>>
Specifies the maximum number of debug trace files allowed. If 0 is specified, there is no limit to the number of files. If the total number of debug trace files exceeds the value specified in this operand, the file with the oldest update date is deleted.
If multiple TP1Client instances for which different values are specified in the dccltdbgtrcfilecount operand are used at the same time, the actual maximum file count cannot be guaranteed.
If you create a TP1Client instance before issuing the rpcOpen method, you can also define the dccltdbgtrcfilecount operand as a system property. In this case, when the rpcOpen method is issued after the TP1Client instance is created, the result obtained by analyzing the content of the rpcOpen method is used to overwrite the file count. When the rpcOpen method is issued, the specification of the dccltdbgtrcfilecount operand defined in the system property becomes invalid.
No error message is issued even when the specification of the dccltdbgtrcfilecount operand is invalid. In this case, the default specification for the dccltdbgtrcfilecount operand takes effect.

dccltrpcmaxmsgsize=maximum-size-of-messages-sent-or-received-by-RPC
~<unsigned integer> ((1~8)) <<1>> (MB)
Specifies the maximum size of messages that can be sent or received by RPCs.
If this operand is omitted or if the default (1) is specified, the maximum size of RPC messages that can be sent or received by the rpcCall method is 1 MB (1048576 bytes). If the maximum allowable value (8) is specified for this operand, the maximum message size is 8 MB (8388608 bytes). You cannot specify a send or receive message size that exceeds the value specified in this operand as an argument of the rpcCall method. If the specified send or receive message size exceeds the value specified in this operand, the rpcCall method returns ErrInvalidArgsException.
When you specify the dccltrpcmaxmsgsize operand, do not use the facilities listed below. Otherwise, an error occurs at the communication target TP1/Server node.

RPC that uses the scheduler direct facility
RPC that specifies the communication target
TP1/Web connection facility

The following table shows what happens when a service request is issued to a TP1/Server other than a TP1/Server (version 06-02 or later) that supports the rpc_max_message_size operand of the system common definition:

Table 5-3 Action of the rpcCall method when 2 or more is specified in the dccltrpcmaxmsgsize operand

Method	Definition that specifies the RPC type	Version of TP1/Server that acts as gateway	Node version^#1
Method	Definition that specifies the RPC type	Version of TP1/Server that acts as gateway	Before 06-02	06-02 or later^#2
rpcCall	dcrapdirect=Y	Before 06-02	N^#3	N^#3
	dcrapdirect=Y	06-02 or later	N^#4	Y
	dcnamuse=Y	Before 06-02	N^#4	Y
	dcnamuse=Y	06-02 or later	N^#4	Y

Legend:: Y: Can be requested.; N: Cannot be requested.

#1: Version of the node where the SPP is running and that was defined in the all_node operand specified for the TP1/Server that acts as a gateway.

#2: When the rpc_max_message_size operand is specified in the system common definition.

#3: When the input data exceeds 1 MB, ErrMessageTooBigException occurs; when the input data is 1 MB or smaller and if the output area size exceeds 1 MB, ErrInvalidArgsException occurs.

#4: ErrNoSuchServiceGroupException occurs.

dcscdhostchange=Y|N ~<<N>>
Specifies whether or not the service request target schedulers specified in the dchost operand of the TP1/Client/J environment definition are distributed to the RPCs.

Y: Distributes the service request target schedulers to the RPCs.

N: Does not distribute the service request target schedulers to the RPCs.

When Y is specified, the following processing occurs:
For the first service required by an RPC, if dchostselect=Y is specified in the TP1/Client/J environment definition, a service request target scheduler specified in the dchost operand of the TP1/Client/J environment definition is randomly selected. If dchostselect=N is specified in the TP1/Client/J environment definition, the first service request target scheduler specified in the dchost operand of the TP1/Client/J environment definition is selected.
For the second and subsequent RPC service requests, the service request target schedulers are selected in the round-robin mode according to the order specified in the dchost operand, beginning with the service request target scheduler previously selected.
dccltloadbalance=Y|N ~<<N>>
Specifies whether or not to store in cache the information on the service request target scheduler with the smallest workload from the information on multiple service request target schedulers obtained from the name server.

Y: Stores the information on the service request target scheduler with the smallest workload in the cache.

N: Does not store the information on the service request target scheduler with the smallest workload in the cache.

When Y is specified, the following processing occurs:
If multiple service request target schedulers are stored in the cache, the first service request target scheduler is randomly selected. For the second and subsequent RPC service requests, the service request target schedulers stored in the cache are selected in the round-robin mode.
For details about workload levels, see the manual OpenTP1 Description.
dccltcachetim=cache-expiration-time
~<unsigned integer> ((0~65535)) <<30>> (seconds)
Specifies how long information on a service request target scheduler stored in the cache remains valid. Information on a service request target scheduler that has reached the cache expiration time is deleted from the cache. The cache is updated when information on the service request target scheduler is subsequently acquired again and stored in the cache.
This operand is valid only when dccltloadbalance=Y is specified in the TP1/Client/J environment definition.
dccltconnecttimeout=maximum-time-for-monitoring-connection-establishment
~<unsigned integer> ((0~65535)) <<0>> (seconds)
Specifies the maximum amount of monitoring time it takes to establish a connection during data transmission.
The value specified in this operand is not the processing time specified for the java.net.Socket.connect method (hereafter referred to as the connect method), but is the maximum amount of monitoring time it takes for the connect method to establish a connection.
If a connection cannot be established because the remote system has not been started, for example, the method issued by a CUP may return an error before the monitoring time specified in this operand elapses. This is because the connection establishment monitoring time by the OS takes precedence over the maximum monitoring time specified in this operand. The connection establishment monitoring time by the OS, the number of times connection establishment requests can be resent, and the resend interval vary depending on the platform. Furthermore, depending on the method or function used, the processing time of the connect method may exceed the value specified in this operand.
If 0 is specified or this operand is omitted, the OS monitors the connection establishment process.
dccltprftrace=Y|N ~<<Y>>
Specifies whether or not a performance analysis trace of the Cosminexus Application Server is collected when TP1/Client/J is run on a Cosminexus Application Server.

Y: Collects a performance analysis trace. However, no performance analysis trace is collected if the PRF daemon has not been started.

N: Does not collect a performance analysis trace.
dccltprfinfosend=Y|N ~<<Y>>
Specifies whether or not identification information in TP1/Client/J is added to the information output to the performance verification trace of OpenTP1 when an RPC that uses the name service or scheduler direct facility is issued to TP1/Server.

Y: Adds identification information inside TP1/Client/J.

N: Does not add identification information inside TP1/Client/J.
dccltnammlthost=Y|N ~<<N>>
Specifies whether or not an RPC that uses the name service is issued to a TP1/Server in the multi-homed host environment.

Y: Issues an RPC that uses the name service to a TP1/Server in the multi-homed host environment. Specify this option when multiple RPC request targets are defined or when there are more than one TP1/Servers in the multi-homed host environment.

N: Does not issue an RPC that uses the name service to a TP1/Server in the multi-homed host environment. An error may occur if an RPC that uses the name service is issued to a TP1/Server in the multi-homed host environment.

For details, see 2.2.9(3) Definition for issuing an RPC to TP1/Server in the multi-homed host environment.
dccltdatacomp=Y|N ~ <<N>>
Specifies whether or not the data compression facility is used.

Y: Uses the data compression facility.

N: Does not use the data compression facility.
dccltconnectinf=terminal-identification-information
Specifies (in EBCDIK codes) the logical terminal name of the DCCM3 logical terminal as the terminal identification information. The terminal identification information is expressed as the value 0x followed by a maximum of 128 hexadecimal characters (0x at the beginning is not included in the 128 characters). You can specify a maximum of 64 bytes (128 characters), where each byte consists of 2 characters. Note that DCCM3 uses only the first 8 bytes and ignores the remainder of the value.
You can change this operand value dynamically by calling the setConnectInformation method. Once the setConnectInformation method has been called, the value specified in this operand is ignored until the rpcOpen method is called again.
The value specified in this operand takes effect when DCCM3 logical terminal host name is specified in the dchost operand and the port number of the DCCM3 logical terminal is specified in the dchost or dcrapport operand in the TP1/Client/J environment definition, and a permanent connection is established with the DCCM3 logical terminal by means of one of the following methods:
- Calling the openConnection method. In the case of the openConnection method with parameters, specify DCCM3 logical terminal host name in the host parameter and the port number of the DCCM3 logical terminal in the port parameter.
- Specifying Y in the dcrapautoconnect operand in the TP1/Client/J environment definition and then call the rpcCall method.
If this operand is omitted, terminal identification information is not sent to the DCCM3 logical terminal. However, if the setConnectInformation method is called, the terminal identification information specified in the setConnectInformation method is sent to the DCCM3 logical terminal when a permanent connection with the DCCM3 logical terminal is established.
dcscdmulti=Y|N ~ <<N>>
Specifies whether or not the multi-scheduler facility is to be used.

Y: Uses the multi-scheduler facility.

N: Does not use the multi-scheduler facility.

When you use the multi-scheduler facility, you can reduce the workload of scheduling by randomly selecting one of the multi-scheduler daemons that have been started.
If you specify Y in this definition, also pay attention to the dchost, dcscdport, and dcscdmulticount TP1/Client/J environment definitions. Note that this definition is not applicable during execution of the rpcCallTo method.
dcscdmulticount=processes-count-for-multi-scheduler-daemon ~ <unsigned integer> ((1-4096)) <<1>>
Specifies the number of processes for the multi-scheduler daemon. This value must be equal to or less than the number of processes specified in the -m option of the scdmulti schedule service definition.
This definition is applicable when dcscdmulti=Y and dcscddirect=Y are specified in the TP1/Client/J environment definition. In such a case, the port number is selected randomly from within the following range of values:
- Minimum value: Port number specified in dchost or dcscdport
- Maximum value: Minimum value + number of processes specified in dcscdmulticount - 1
If the maximum value for this port number exceeds 65535, ErrFatalException occurs during definition analysis and ErrInvalidPortException occurs during setDchost processing.
If the specified value is greater than the number of processes specified in the -m option of scdmulti, an attempt may be made to establish connection with a port where the multi-scheduler is not running, resulting in the ErrNetDownAtClientException.
dccltcupsndhost=transmission-source-host ~ <character string>
Specifies the transmission source host during the following connection establishments:
- RPC using the scheduler direct facility
- RPC using the name service
- RPC using the remote API facility
- RPC specifying the communication target
- TCP/IP communication facility
For the host name, you can also specify an IP address in decimal dot notation.
An exception occurs on the issued method in the following cases:
- localhost was specified as the host name
- The specified IP address begins with 127
- The specified host does not exist on the system where the CUP is being executed
If this definition is omitted, the transmission source host is allocated arbitrarily.
dccltcuprcvport=port-number-used-for-reception-of-CUP ~ <unsigned integer> ((5001-65535))
Specifies the port number of the CUP that is to receive messages from the server.
The port number specified in this definition takes effect when the following functions are used:
- When an RPC using the scheduler direct facility is received (receive port)
- When an RPC using the name service is received (receive port)
If this definition is omitted, a port number allocated arbitrarily by the system is used.
If you execute multiple processes or threads concurrently within the same machine, specify different port numbers. If a duplicated port number is specified, correct reply data may not be received. Note that different OSs use different port number; for details, see the manual for the applicable OS.
dcclttrcmplmttm=maximum-time-allowed-for-completion-of-transaction ~ <unsigned integer> ((0-65535)) (seconds)
Specifies the maximum execution time from startup to termination of a transaction branch that is executed by the RAP-processing server. If the specified time is exceeded, the transaction branch settles to either rollback or commit by the recovery process, and then is terminated after the RAP-processing server processing terminates abnormally.
If 0 is specified, the maximum transaction branch execution time is not monitored.
If this specification is omitted, processing depends on the trn_completion_limit_time operand specification in the RAP-processing listener service definition.
If this facility is used with a TP1/Server version earlier than 07-01, the facility runs in invalid status.

5.2.2 Operands