- The MAX control statement allows you to define the maximum values, timers, process values and system parameters of a UTM application. For instance, these include:
the name of the application
the base name or the base directory for UTM files
single or dual-file operation of the KDCFILE
size of a UTM page (block size of UTM storages and buffers)
the maximum number of
processes
key codes
GSSBs
LSSBs
UTM pages in the buffer for user log records, etc.
threshold values for monitoring the size of SYSLOG file generations if the SYSLOG is created as an FGG (SYSLOG-SIZE operand)
the default language environment of the UTM application (LOCALE operand)
whether or not SM2 can be used for performance monitoring in the application
The parameters of the MAX control statement can be split into several MAX statements. If the same operand is inadvertently entered in several MAX statements, the first value entered for this operand is taken as valid.
Mandatory operands:
APPLINAME=, KDCFILE= and TASKS=.
Additional mandatory operands on Unix, Linux and Windows systems:
SEMKEY= or SEMARRAY= (semaphore keys), IPCSHMKEY=, KAASHMKEY= and CACHESHMKEY=.
In OSI TP applications additionally: XAPTPSHMKEY= and OSISHMKEY=.
The mandatory operands need to be defined once.
Note on UTM cluster applications on Unix, Linux and Windows systems:
If you modify one of the operands APPLINAME, APPLIMODE, GSSBS, LSSBS, KB or NB then you must regenerate both the initial KDCFILE and the UTM cluster files by specifying GEN=(CLUSTER,KDCFILE) in the OPTION statement.
For clarity, all operands of the MAX statement are listed in a table following the operand descriptions.
Operands valid for all operating systems
| [ ,PGPOOL=( number,warnlevel1,warnlevel2 ) ] [ ,PGPOOLFS=number ] [ ,PGWTTIME=time ] [ ,PRIVILEGED-LTERM = <lterm-name> ] [ ,QTIME = (qtime1,qtime2)] [ ,RECBUF=( number,length ) ] [ ,RECBUFFS=number ] [ ,REDELIVERY=(number1, number2) ] [ ,RESWAIT={ time1 | ( time1, time2 ) } ] [ ,SM2={ NO | OFF | ON } ] [ ,SPAB=length ] [ ,STATISTICS-MSG={ NONE | FULL-HOUR } ] [ ,SYSLOG-SIZE=size ] [ ,SYSTEM-TASKS={ *STD | number } ] ,TASKS=number [ ,TASKS-IN-PGWT=number ] [ ,TERMWAIT=time ] [ ,TRACEREC=number ] [ ,TRMSGLTH=length ] [ ,USLOG={ SINGLE | DOUBLE } ]further operands for BS2000 systems
further operands for Unix, Linux and Windows systems
|
1NORES | RES and PS | DS only permitted on BS2000 systems. The long form PROGRAM-SPACE or DATA-SPACE can also be specified instead of PS or DS.
APPLIMODE= This specifies whether the application is a UTM-S or UTM-F application. | ||||||||||||||||||||||
SECURE | The application is generated as a UTM-S application. With UTM-S, openUTM logs all user data so that this data is retained after the application is terminated or following a system crash. In the event of errors, UTM-S guarantees the integrity and consistency of the application data. If a UTM-S application is terminated abnormally, an automatic restart is automatically performed. For this purpose, this variant of openUTM logs all modifications at the end of transactions. Default: SECURE | |||||||||||||||||||||
FAST | The application is generated as a UTM-F application. UTM-F offers enhanced performance by eliminating the disk input/output operations performed by UTM-S when logging user and transaction data. With a standalone UTM-F application, openUTM only logs user passwords and changes to the configuration which were made by means of dynamic administration. These modifications are thus retained for the next application run. However, UTM-F applications do not log changes to the user data. They are therefore suitable only for installations in which performance is the most important criterion and the restart facility is not required. This applies in the case of pure information systems, or if all logging functions can be provided by the database system used. In UTM cluster applications, user data that is valid globally in the cluster is also saved for UTM-F. | |||||||||||||||||||||
APPLINAME= | appliname Name of the UTM application up to eight characters in length. appliname defines a transport system access point via which connections to the UTM application can be established. This is a mandatory operand. If several application names are required, for example, for distributed processing based on the LU6.1 protocol, these can be assigned to the application using the BCAMAPPL statement. With APPLINAME= you define the primary application name. appliname must be unique within the local system and must not begin with the character ’$’. BS2000 systems: This name is subject to the name conventions for BCAM applications. The transport system access point defined with appliname supports the transport protocol NEA. appliname must not begin with a number or with ’$’ as this is prohibited by BCAM and the application cannot be started otherwise. Please note that KDCDEF cannot intercept numbers. Unix, Linux and Windows systems: appliname must be specified when establishing a connection from the terminal (dialog terminal process). If connections are to be established with partner applications using the application name defined with APPLINAME=, you must also issue an appropriate BCAMAPPL statement (see "BCAMAPPL - define additional application names"). | |||||||||||||||||||||
ASYNTASKS= | (atask_number,service_number) Maximum number of resources that may be reserved to process asynchronous jobs. | |||||||||||||||||||||
atask_number | Maximum number of processes (BS2000 tasks or work processes on Unix, Linux and Windows systems) of the application which can simultaneously handle jobs with asynchronous transaction codes. This operand allows you to prevent long-running asynchronous processes from affecting dialog operation. If ASYNTASKS=0, asynchronous TAC classes cannot be generated. Default: 1 | |||||||||||||||||||||
service_number | Maximum number of asynchronous services that may be open at the same time. You should set service_number to be larger than atask_number when one of the two following cases can arise:
If these cases arise in the application and the value of service_number is too small, then the asynchronous processing may be temporarily blocked because service_number of inactive services already exist. New asynchronous services cannot be started although no UTM processes are processing asynchronous services at this time. Default: atask_number | |||||||||||||||||||||
BLKSIZE= | Size of a UTM page Please note that, depending on the BLKSIZE specification, each user storage area occupies at least 2K, 4K or 8K in the page pool. You can only specify BLKSIZE=4K or 8K for UTM cluster applications. Default
Possible values: 2K, 4K, 8K On BS2000 systems you must specify BLKSIZE=4K or 8K
| |||||||||||||||||||||
BRETRYNR= | number This operand is only supported on BS2000 systems. Number of attempts made by openUTM to transfer a message to the transport system (BCAM) if BCAM cannot accept the message immediately. If this number is exceeded, the connection to the dialog partner is shut down. BRETRYNR is irrelevant for asynchronous messages output to a dialog partner with PTYPE=APPLI (PTERM statement). If such a message from the transport system is rejected due to a temporary bottleneck, then openUTM releases the process, but does not clear down the connection. After waiting for three seconds, openUTM makes up to three attempts to transfer the message to BCAM. If after the third attempt BCAM still cannot accept the message, then openUTM waits for 3 more seconds before it makes another three attempts to send the message to BCAM. If still unsuccessful, it waits another 3 seconds before making another three attempts, and so on. Default: 10 | |||||||||||||||||||||
CACHESHMKEY= | number This operand is only supported on Unix, Linux and Windows systems. Authorization key for the shared memory segment containing the global buffer for file access. Keys are global parameters under the Unix, Linux and Windows systems. You cannot specify more than one key. You must enter a decimal number for number. This is a mandatory operand. | |||||||||||||||||||||
CACHESIZE= | (number,paging,NORES or RES, PS or DS) (NORES, RES, PS, DS only on BS2000 systems PROGRAM-SPACE or DATA-SPACE can also be specified instead of PS or DS) This specifies the size and properties of the cache memory (further information can be found in the openUTM manual “Concepts und Functions”). The values entered here affect the performance of your UTM application. | |||||||||||||||||||||
number | Number of UTM pages in the cache. The size of each UTM page is defined in the BLKSIZE= operand. The cache is used for accessing the page pool, i.e. all input and output operations involving LSSBs, GSSBs, TLSs, LPUT messages, FPUT messages, MPUT messages to clients, and some types of UTM administrative data. Data is not written to the KDCFILE until the cache becomes full or the transaction is terminated. KDCDEF rounds up this number to a multiple of 32. Default value: If the cache on BS2000 systems is located in the program space (PS), the cache is created in a common memory pool whose size is always a multiple of 1 MB. The BS2000 system automatically rounds the value specified in CACHESIZE. CACHESIZE should be requested in multiples of 1 MB so that address space is not wasted. | |||||||||||||||||||||
paging | Percentage of cache pages to be written to the KDCFILE in a single batch in the event of a bottleneck, thereby freeing space in the cache. This must correspond to at least eight pages. The value specified here can be modified using the administration command KDCAPPL CACHE=%_utm_pages. Default value: 70(%) | |||||||||||||||||||||
NORES | This operand value is only supported on BS2000 systems. The cache is created as non-resident. Default: NORES | |||||||||||||||||||||
RES | This operand value is only supported on BS2000 systems. The cache is created as resident. A resident cache can enhance the performance of the UTM application. RES may not be specified together with DATA-SPACE. Resident cache offers enhanced performance in productive mode, as the cache paging algorithm is designed specifically for use with this type of cache. The number of resident pages used in the creation of a resident cache cannot be checked using the COREBIAS operand of the BS2000 command BIAS. | |||||||||||||||||||||
PS or PROGRAM-SPACE | This operand value is only supported on BS2000 systems. The UTM cache is created in the program space. Default: PS | |||||||||||||||||||||
DS or DATA-SPACE | This operand value is only supported on BS2000 systems. The UTM cache is created in one or more data spaces. If the generated UTM cache is larger than 2GBt, UTM will distribute the cache over more than one data space as a data space may be at most 2GB in size. The option of creating the UTM cache in a data space should be chosen only if a very large UTM cache is required and the address space (program space) is not sufficient for this. Use of a data space for the ITM cache always entails a slight loss of performance as a result of the way in which a program can access data in a data space. For applications which require a large UTM cache, these performance disadvantages are, however, counterbalanced by the advantages which a large cache brings through the reduction of file IOs. Above all for UTM applications it can be advantageous to create a very large UTM cache in a data space. In the case of UTM-F, cache buffers are written to file only in the event of a cache bottleneck. If the UTM cache is generated large enough all file IOs to this page pool may be omitted for such applications. The maximum size of a UTM cache in data spaces is 8 GB. In other words:
DATA-SPACE may not be specified together with RES. | |||||||||||||||||||||
CARDLTH= | length This operand is only supported on BS2000 systems. Length of the ID card information in bytes. If the ID card reader is used for sign-on, openUTM stores the ID card information in the length resulting from the maximum of the length specified here and the value generated for MAX PRINCIPAL-LTH. If the information on the ID card is longer, it is truncated and stored in this length. The KDCS call INFO (KCOM=CD) enables a program to read this information. CARDLTH must be big enough to ensure that the following applies for all USER statements with Default: 0 When a value > 255 is specified, 255 is assumed. | |||||||||||||||||||||
CATID= | (catalog_A,catalog_B) This operand is only supported on BS2000 systems. Catalog IDs to which your KDCFILE is assigned. If you work with CATIDs, enter the base name without the CATID in KDCFILE=filebase (see "MAX - define UTM application parameters"). In the case of single-file operation of the KDCFILE, specify the CATID to which the KDCFILE is to be assigned in catalog_A. In this case, catalog_B is not specified. In the case of dual-file operation of the KDCFILE (see "The KDCFILE"), you can assign files with the suffix A to CATID catalog_A and files with the suffix B to catalog_B. If you only specify a value for catalog_A, both files are assigned to this CATID. | |||||||||||||||||||||
CLRCH= | Character with which the KB program area and the standard primary working area are overwritten at the end of a dialog step. Possible entries are: c Where c is an alphanumeric character and x a hexadecimal character. Default: | |||||||||||||||||||||
CONN-USERS= | number This operand is used to control the load on the application. It defines the maximum number of users that can work simultaneously with the application. In the case of an application for which user IDs have not been generated, CONN-USERS= can be used to define the maximum number of clients that can sign simultaneously on to the application via LTERM partners.
User IDs and clients generated with administration authorization can sign on to the UTM application, even if the maximum number of simultaneously active user IDs has already been reached. Default value on BS2000 systems: 0 (i.e. no restriction) CONN-USERS is a mandatory operand on Unix, Linux and Windows systems. Please note that number cannot be set to a higher value that the number of concurrent user licenses obtained. | |||||||||||||||||||||
CONRTIME= | time (connection request time) Time in minutes after which openUTM retries to establish a connection after failing to establish a connection generated to be established automatically.
If a connection to this partner is not established when the application is started or the administration command KDCPTERM or KDCLPAP is issued, openUTM attempts to reestablish the connection at intervals specified in CONRTIME=. If CONRTIME=0, openUTM does not make any attempt to set up the connection. Exception: A wait time of 10 minutes is set for asynchronous messages to OSI TP partners. Default: 10 min. | |||||||||||||||||||||
DATA-COMPRESSION= | This parameter enables data compression to be permitted or not permitted. | |||||||||||||||||||||
YES | Data compression is permitted and enabled. Data compression can be disabled using administration facilities, e.g. by means of KDCAPPL. | |||||||||||||||||||||
NO | Data compression is not permitted and is disabled. This setting cannot be modified using administration facilities. | |||||||||||||||||||||
STD | Data compression is permitted and enabled for UTM-S applications (APPLIMODE=S). This default setting can be modified using administration facilities. Default: STD The average value for UTM pages saved per data compression can be queried using the administration functions, e.g. using the KDCINF STAT command (see openUTM manual “Administering Applications”) or using WinAdmin or WebAdmin. | |||||||||||||||||||||
DEAD-LETTER-Q-ALARM | Controls monitoring the number of messages in the dead letter queue. The K134 message is output each time the threshold is reached. for this message the destination MSGTAC can be defined in order to automate handling of the dead letter queue. Default: 0, monitoring is disabled. | |||||||||||||||||||||
DESTADM= | destination Destination to which openUTM sends the results of administration calls processed asynchronously. For destination, you can specify:
Default: Blanks, i.e. no destination; the results are thus lost. | |||||||||||||||||||||
DPUTLIMIT1= | (day,hour,minute,second) Defines the latest possible execution time of a job. Can be specified in relative or absolute time: time of execution < time of DPUT call + DPUTLIMIT1 The following applies for time specifications in DPUTLIMIT1: | |||||||||||||||||||||
day | Maximum value: 364 | |||||||||||||||||||||
hour | Maximum value: 23 | |||||||||||||||||||||
minute | Maximum value: 59 | |||||||||||||||||||||
second | Maximum value: 59 Default value: DPUTLIMIT1 (360, 0, 0, 0) The following must apply for the DPUTLIMIT1 and DPUTLIMIT2 operands: DPUTLIMIT1 + DPUTLIMIT2 <= (364, 23, 59, 59) < 365 days i.e. if you enter (364, 23, 59, 59) for DPUTLIMIT1, you must specify DPUTLIMIT2=(0, 0, 0, 0). | |||||||||||||||||||||
DPUTLIMIT2= | (day,hour,minute,second) The time specification for the DPUT call does not contain a number for the year. Furthermore, the desired execution time may already have passed if the DPUT call was delayed. For this reason, you must decide whether the execution time of a job with an absolute time specification should be attributed to the past, current, or next year. Since DPUTLIMIT1 + DPUTLIMIT2 must be < 1 year, only one of these three alternatives will be in the permissible open time period (call time - DPUTLIMIT2, call time + DPUTLIMIT1):
DPUTLIMIT2 therefore allows you to backdate the specified execution time into the past for time-driven jobs with absolute time specifications. You cannot backdate jobs with relative time specifications. DPUTLIMIT1 restricts the predating of jobs with absolute or relative time specifications into the future only. Example 1
The DPUT call time is (005,0,0,0). The current and last years are not leap years.
DPUTs with absolute time (305,0,0,0) to (360,0,0,1) are rejected. -|---------------|------------- ... --------------|-----------| 360 5 305 360 FFFFFFFFFFFFFFFFFDDDDDDDDDDDDD ... DDDDDDDDDDDDDDD Example 2
-|---------------|-------------- ... -----------|-------------| 350 360 295 350 FFFFFFFFFFFFFFFFFDDDDDDDDDDDDDD ... DDDDDDDDDDDD The default values are listed under the description of the DPUTLIMIT1 operand. | |||||||||||||||||||||
GSSBS= | number Maximum number of GSSBs (global secondary storage areas) that can exist simultaneously in the application. Default: 32 | |||||||||||||||||||||
HOSTNAME= | name BS2000 systems: Name of the virtual host on which the UTM application runs (from the point of view of BCAM). This virtual host must also be generated in BCAM. The name can be up to 8 characters in length. Default value: 8 blanks, i.e the applications runs under the real host. Unix, Linux and Windows systems: Can only be specified in standalone applications. In UTM cluster applications, you can specify a virtual host name in the VIRTUAL-HOST parameter of the CLUSTER-NODE statement. Name of the host that is specified as the sending address when a connection is established from the UTM application end. HOSTNAME= is required in cluster systems that use the “relocatable” IP address as the sending address and not the stationary IP address. Default: Blanks, the default processor name of the transport system is used as the sending address. | |||||||||||||||||||||
IPCSHMKEY= | number This operand is only supported on Unix, Linux and Windows systems. Authorization key for the shared memory segment, which is used for communication between work processes on one side and the dialog terminal or printer processes and the timer process (external processes of an application) on the other side. Keys are global parameters under the Unix, Linux and Windows systems. You cannot specify more than one key. You must enter a decimal number for number. This is a mandatory operand. | |||||||||||||||||||||
IPCTRACE= | number This operand is only supported on Unix, Linux and Windows systems. In test mode (startup with TESTMODE=ON, see openUTM manual “Using UTM Applications on Unix, Linux and Windows Systems”), openUTM writes entries in the trace area of the IPC (shared memory segments for interprocess communication). These entries contain internal information which is required for diagnostic purposes. Each entry occupies 32 bytes. IPCTRACE defines the number of entries in the trace area. If this number is exceeded, openUTM overwrites the existing entries, beginning with the oldest entry. Default: 1060 KDCDEF automatically resets values < 1 or > 32500 to the minimum or maximum value without outputting a UTM message. | |||||||||||||||||||||
KAASHMKEY= | number This operand is only supported on Unix, Linux and Windows systems. Authorization key for the shared memory segment containing the global data. Keys are global parameters under the Unix, Linux and Windows systems. You cannot specify more than one key. You must enter a decimal number for number. This is a mandatory operand. | |||||||||||||||||||||
KB= | length Length of the communication area (KB) in bytes, excluding the KB header and KB return area. Default: 512 | |||||||||||||||||||||
KDCFILE= | ||||||||||||||||||||||
filebase | Base name of the KDCFILE, the user log file, and the system log file SYSLOG. The name entered here must also be specified in the start parameter FILEBASE=filebase when starting the application program (see openUTM manual “Using UTM Applications”). This is a mandatory operand. BS2000 systems: Unix, Linux and Windows systems: | |||||||||||||||||||||
SINGLE | Single-file operation is activated for the KDCFILE. If the KDCFILE is split (see section "Splitting the KDCFILE"), all KDCFILE files are subject to single-file operation. Default: SINGLE | |||||||||||||||||||||
DOUBLE | For security reasons, dual-file operation is activated for the KDCFILE. If the KDCFILE is split (see section "Splitting the KDCFILE"), all KDCFILE files are subject to dual-file operation. In UTM cluster applications, only SINGLE may be specified. | |||||||||||||||||||||
KEYVALUE= | number Value of the highest key code of the application, and thus the value of the corresponding highest lock code that can be assigned to a transaction code or a terminal for data access control. The operand KEYVALUE=number can also be used to define the maximum number of key codes per key set. openUTM uses this information to optimize the key set tables. You can define up to 4000 key and lock codes. Only numerical lock codes can be defined. Default: 32 Exceptions:
If you enter a value < 1, KDCDEF automatically sets KEYVALUE=1 without outputting a UTM message. | |||||||||||||||||||||
LEADING-SPACES= | Specifies how the leading spaces in a messages from a terminal or from a TS application (PTERM ... PTYPE=APPLI or SOCKET) are to be handled. | |||||||||||||||||||||
YES | When calling a program unit, leading blanks in messages are passed on to the program unit. The same applies for messages sent to a client with PTYPE=APPLI. A blank acting as a separator between TAC and message is removed if the TAC name < 8 characters. | |||||||||||||||||||||
NO | Leading blanks are suppressed. Default: NO | |||||||||||||||||||||
LOCALE= | (lang_id,terr_id,ccsname) This operand is only supported on BS2000 systems. Default language environment of the UTM application (see also section"UTM messages"). The locale generated here is assigned to all user IDs and clients that sign on via LTERM partners or LTERM pools as the default language environment. This default setting applies unless another locale is explicitly defined for these objects in the corresponding USER, LTERM, or TPOOL statements. The message module whose language and territorial identifiers match the specifications in the MESSAGE ...LOCALE= and MAX ...,LOCALE= statements becomes the application message module. openUTM sends messages to the message destinations SYSOUT, SYSLST, and CONSOLE from this application message module. The specifications in the application message module also determine the destination of a particular message. | |||||||||||||||||||||
lang_id | Freely selectable language identifier for the UTM application up to two characters in length. Default: Blanks | |||||||||||||||||||||
terr_id | Freely selectable territorial identifier up to two characters in length. Default: Blanks | |||||||||||||||||||||
ccsname | (coded character set name) Default: Blanks, i.e. 7-bit mode | |||||||||||||||||||||
LOGACKWAIT= | time The maximum length of time in seconds that openUTM is to wait for an acknowledgment from an output device. This acknowledgment is
If confirmation does not arrive within this period, e.g. because the printer has run out of paper, openUTM shuts down the logical connection to the device. Default: 600 | |||||||||||||||||||||
LPUTBUF= | number Size of the LPUT buffer in UTM pages. The LPUT buffer of the KDCFILE is used to temporarily store LPUT data. This data is not copied to the user log file until the value specified in number is exceeded. The user log file USLOG is open only during this copy process. Default: 1 CAUTION! LPUTBUF * UTM page size >= LPUTLTH + length of KB header (84 bytes) | |||||||||||||||||||||
LPUTLTH= | length Maximum length of the user data in LPUT records in bytes (excluding the KB header). The maximum length of an LPUT record in the user log file is calculated as follows (see also the openUTM manual „Programming Applications with KDCS”, user log file): length + 84 bytes for the KB header + 12 bytes for length fields. Default: 1948 BS2000 systems: If the user log file USLOG is created on a non-key disk (NK2, NK4), then you must select the value of length such that: length + 100 byte + 16 bytes block-specific internal DVS administration information is a multiple of 2 Kbytes (on NK2 disks) or 4 Kbytes (on NK4 disks). This allows you to optimally utilize disk space. The 16 byte block-specific internal DVS administration information are therefore not available for use as user data. You will find more information on this subject in the BS2000 manual "Introductory Guide to DMS". | |||||||||||||||||||||
LSSBS= | number Maximum number of LSSBs (local secondary storage areas) that can be created in a service. Default: 8 | |||||||||||||||||||||
MOVE-BUNDLE-MSGS= | This parameter can be used for an application to allow automatic moving of waiting asynchronous messages from a slave LTERM, slave LPAP or slave OSI-LPAP without a connection to the partner application. | |||||||||||||||||||||
YES | When the waiting time defined in MAX CONRTIME has elapsed, or after 10 | |||||||||||||||||||||
NO | Asynchronous messages to a slave are never sent via another slave. Default: NO | |||||||||||||||||||||
MP-WAIT= | number This operand is only supported on BS2000 systems. Maximum number of seconds for which openUTM waits for a process to sign on to a common memory pool. Default value: 180 CAUTION! The default value of 180 seconds should only be changed in exceptional circumstances, e.g. if a process terminates with K078 ENQAR and a user dump with the return code KDCSST01. | |||||||||||||||||||||
NB= | length Maximum length of a working area for
This should be equal to the length of the largest KDCS message area of the program units in bytes. Default: 2048 | |||||||||||||||||||||
NRCONV= | number (number of conversations) The following limits are valid: Number of user IDs + maximum number of services that can be placed on the stack (number of services =number * number of user IDs) <= 500000 If the limit value of 500000 is exceeded (by the values specified for NRCONV in the RESERVE statement, see "RESERVE - reserve table locations for UTM objects", and by the number of USER statements, see "USER - define a user ID"), then openUTM automatically creates fewer entries for stacking services. In this case, not all users will be able to place number services on the stack. Default: 0 | |||||||||||||||||||||
OSISHMKEY= | number This operand is only supported on Unix, Linux and Windows systems. Authorization key for the shared memory segment, which is used by OSS for communication based on OSI TP. You must enter a decimal number for number. | |||||||||||||||||||||
OSI-SCRATCH-AREA= | value Size in KB of an internal UTM working area for dynamic data storage when using the OSI TP protocol. Default: 256 On BS2000 systems this working area is automatically extended during runtime, if required. On Unix, Linux and Windows systems the size of the internal working area must not be modified during runtime. It is recommended that you select the default value. However, if this proves to be insufficient during operation, increase the value of OSI-SCRATCH-AREA and repeat the generation procedure. | |||||||||||||||||||||
PGPOOL= | (number,warnlevel1,warnlevel2) Size of the page pool in UTM pages and the warning levels for utilization of the page pool. | |||||||||||||||||||||
number | Number of UTM pages to be used for the page pool in the KDCFILE (see "Page pool"). The size of each UTM page is defined in the BLKSIZE= operand. Default: 100 If you enter a value less than 20, KDCDEF automatically sets PGPOOL=20 without outputting a UTM message. On Unix, Linux and Windows systems the value of PGPOOL is always an even number. If you enter an uneven number, openUTM subtracts 1 from your entry. | |||||||||||||||||||||
warnlevel1 | Numeric value (percentage) indicating the page pool utilization level at which the first warning (UTM message K041) is output. Default: 80 | |||||||||||||||||||||
warnlevel2 | Numeric value (percentage) indicating the page pool utilization level at which the second warning is to be output. If warnlevel2 is exceeded, all asynchronous jobs are rejected. In this case, the user receives a K message, and a corresponding return code is sent to a program unit. Default: 95 | |||||||||||||||||||||
PGPOOLFS= | number Number of files between which the page pool is to be split. If PGPOOLFS = 0, the page pool is located in the main file (on BS2000 systems in the file filebase.KDCA, on Unix, Linux and Windows systems in the file KDCA in the filebase file directory). In the case of dual-file operation (MAX ...,KDCFILE=(...,DOUBLE)), the value specified in number does not include the two file copies. Default: 0, i.e. the page pool is located in the main file Maximum value (BS2000 systems): 99 (and PGPOOL=number / 2) Minimum value: The minimum value depends on the number of UTM pages, the UTM page size and the maximum file size permitted on the relevant system.
| |||||||||||||||||||||
PGWTTIME= | time Maximum number of seconds for which a program unit can wait for messages to arrive after a blocking call (e.g. PGWT call). During this period, a process of the UTM application is exclusively reserved for this program unit. Default: corresponds to time in TERMWAIT=time | |||||||||||||||||||||
PRINCIPAL-LTH= | length This operand is only supported on BS2000 systems. Maximum length of a Kerberos principal in bytes. This parameter is only of significance if at least one user is generated with USER ..., PRINCIPAL= or at least one LTERM or TPOOL is generated with KERBEROS-DIALOG=YES. The length of the value specified with USER ... PRINCIPAL= must not be larger than the value generated with MAX PRINCIPAL-LTH=. When a Kerberos dialog is performed with a client, openUTM saves the Kerberos information in the length resulting from the maximum of this length and the length generated for MAX CARDLTH. If the Kerberos information is longer, it is truncated to this length and stored. Default: 0 | |||||||||||||||||||||
PRIVILEGED-LTERM=
| lterm-name Identifies an LTERM as a privileged connection. Jobs sent to the UTM application via this LTERM are prioritized for processing by UTM in situations in which the UTM application is subject to a high load. To permit rapid responsiveness even in high-load situations, additional processes (referred to as UTM system processes) are started for a UTM application. The UTM system processes only handle selected jobs. These are primarily internal jobs or jobs issued by an administrator who is signed on at the UTM application via the privileged LTERM. See also operand MAX SYSTEM-TASKS on "MAX - define UTM application parameters". If optimum use is to be made of this functionality, the PRIVILEGED-LTERM should always be explicitly generated. Only then is it possible for all the mechanisms that allow this LTERM to be privileged in high-load situations to take effect. More specifically, the following approach is recommended:
If a connection is established via this LTERM then the following applies:
The LTERM must be generated as a dialog LTERM in an LTERM statement. If no PRIVILEGED-LTERM is generated then it is dynamically determined as follows:
| |||||||||||||||||||||
QTIME= | (qtime1, qtime2) Specifies the maximum permitted length of time that a service is to wait for the arrival of a message in a message queue. QTIME= refers to user specific (USER queues), permanent (TAC queues) and temporary message queues. It is possible to define individual maximum values for wait times in dialog or asynchronous services. | |||||||||||||||||||||
qtime1 | Maximum length of wait time for dialog services | |||||||||||||||||||||
qtime2 | Maximum length of wait time for asynchronous services Both times are specified in seconds. Default: 32767 (seconds) | |||||||||||||||||||||
RECBUF= | (number,length) Size of the transaction-oriented restart area. This area contains the data required for a restart following a transaction or system error. Further information on the restart area can be found on section "Restart area". | |||||||||||||||||||||
number | Number of UTM pages per process to be used in the KDCFILE to store data for a restart following a system error. The size of each UTM page is defined in the BLKSIZE= operand. If this area is large, the application load is reduced but the restart process following a system error is slower. If this area is small, the application load is increased but the restart process following a system error is faster. Default: 100 (per process) | |||||||||||||||||||||
length | Size in bytes of the buffer available to each application process for temporarily storing restart data. This data is required for a restart following a transaction or system error. Default: 8192 | |||||||||||||||||||||
RECBUFFS= | number Number of files between which the restart area is to be split. If RECBUFFS=0, the restart area is located in the main file of KDCFILE. In the case of dual-file operation (MAX ..., KDCFILE=(...,DOUBLE)), the value specified in number does not include the two file copies. The file names are defined by KDCDEF. number must not be greater than the maximum number of processes defined in TASKS=. If this requirement is not fulfilled, the default value is used. Default: 0 | |||||||||||||||||||||
REDELIVERY= | (number1, number2) Maximum number of redeliveries of an asynchronous message after the service or transaction was rolled back. number1 and number2 apply for different message destinations. | |||||||||||||||||||||
number1 | Maximum number of redeliveries of messages to an asynchronous TAC. Delivery is always repeated after an asynchronous service was terminated abnormally with PEND ER/FR or system PEND ER without at least one transaction having been completed successfully. Restart of an asynchronous service after PEND RS within the first transaction is not regarded as a redelivery. | |||||||||||||||||||||
number2 | Maximum number of redeliveries of messages to a service controlled queue. Delivery is always repeated if the message was processed and the transaction was then rolled back. With the DGET call, the number of redeliveries is output in the KB return area. Default: (0, 255) A value of 0 means that the message is deleted or saved to the dead letter queue after rollback, depending on the value in TAC ...,DEAD-LETTER-Q. If the value is set to 255, a message is redelivered any number of times. Note that this can result in an endless loop if, for example, a program unit is rolled back because of a programming error. Additionally the message cannot be saved to the dead letter queue in case of an endless loop. | |||||||||||||||||||||
REQNR= | number This operand is only supported on BS2000 systems. Maximum number of PAM read/write jobs that can be issued in parallel at the same time for a file in a UTM process. This value can be used to control the parallel processing of input/output operations within certain limits. Default: 20 | |||||||||||||||||||||
RESWAIT= | (time1,time2) (resource wait) | |||||||||||||||||||||
time1 | Maximum number of seconds for which a program unit can wait for a resource locked by another transaction: GSSBs, TLSs, ULSs, and on BS2000 systems possibly LTERM partners if ANNOAMSG=N. RESWAIT=0: The application program does not wait for the resource to become available. If the resource is locked by another transaction, the requesting program unit immediately receives an appropriate return code. Default: 120 On BS2000 systems the real waiting time depends on the precision with which the bourse waiting time was set in the operating system. | |||||||||||||||||||||
time2 | Maximum number of seconds for which you can wait for a resource locked by another process. If time2 is exceeded, the application is terminated abnormally. time2 should not be set too low, since certain activities in the UTM application must be performed and completed by a process before the same activities can be initiated in another process. Example When sending a message, a process locks the terminal to which the message is directed. If another process wishes to access an input message of the same terminal, it must wait for the terminal to become available again. In particular, the value entered for time2 must be at least equal to the longest processing time (real time) required in the following cases:
Default: 300 If the value 0 is specified for time2, KDCDEF uses the default value 300 without outputting a UTM message. If you specify a value between 0 and 300, however, KDCDEF issues an appropriate UTM message. | |||||||||||||||||||||
SAT= | (security audit trail) This operand is only supported on BS2000 systems. Minimum event logging with SAT. Further information about "SAT logging" can be found in the openUTM manual “Using UTM Applications on BS2000 Systems”. | |||||||||||||||||||||
ON | SAT logging is switched on. Minimum logging with SAT is switched on for the following events:
Minimum logging can be extended and controlled by means of preselection. | |||||||||||||||||||||
OFF | SAT logging is switched off Default: OFF | |||||||||||||||||||||
SEMARRAY= | (number,number1) This operand is only supported on Unix, Linux and Windows systems. Range of semaphore keys for global semaphores (process synchronization). Semaphore keys are global parameters under the Unix, Linux and Windows systems. With SEMARRAY, you enter an initial value number and an upper limit number. openUTM then reserves these keys, incrementing them by 1 starting with the initial value. For further information, please contact your system administrator. This is a mandatory operand if SEMKEY= is not specified. The SEMARRAY= and SEMKEY= parameters are mutually exclusive. Compared to SEMKEY=, SEMARRAY= offers the advantage of allowing openUTM to reserve more than ten semaphore keys. To calculate the number of semaphore keys required for a UTM application, please refer to the description of the global system resources in the openUTM manual “Using UTM Applications on Unix, Linux and Windows Systems”. | |||||||||||||||||||||
number | Initial value (numeric value) | |||||||||||||||||||||
number1 | Number of keys to be reserved Minimum value: 1 | |||||||||||||||||||||
SEMKEY= | (number,...) This operand is only supported on Unix, Linux and Windows systems. This is a mandatory operand if SEMARRAY is not specified. The SEMARRAY= and SEMKEY= parameters are mutually exclusive. Compared to SEMKEY=, SEMARRAY= offers the advantage of allowing openUTM to reserve more than ten semaphore keys. To calculate the number of semaphore keys required for a UTM application, please refer to the description of the global system resources in the openUTM manual “Using UTM Applications on Unix, Linux and Windows Systems”. | |||||||||||||||||||||
SM2= | This defines whether the UTM application is to supply data to SM2 or openSM2 for performance monitoring. | |||||||||||||||||||||
NO | Performance monitoring with openSM2 is generally prohibited for the UTM application, i.e. the UTM application cannot supply data to openSM2, nor can this be explicitly activated by the UTM administrator. | |||||||||||||||||||||
OFF | The UTM application can supply data to openSM2, but this must be explicitly activated by the administrator using KDCAPPL SM2=ON. The supply of data can be deactivated again at any time using the administration command KDCAPPL SM2=OFF. Default value: OFF | |||||||||||||||||||||
ON | The UTM application can supply data to openSM2. This is activated automatically when starting the UTM application. It can be deactivated again at any time by the UTM application administrator using the administration command KDCAPPL SM2=OFF. | |||||||||||||||||||||
SPAB= | length Maximum length of the standard primary working area in bytes Default: 512 | |||||||||||||||||||||
STATISTICS-MSG= | Specifies whether or not openUTM is to produce statistics message K081 hourly. | |||||||||||||||||||||
FULL-HOUR | Statistics message K081 is produced every hour and written in the SYSLOG. At the same time, openUTM resets the following application specific statistical values to 0:
| |||||||||||||||||||||
NONE | Statistics message K081 is not produced and the statistical values listed above are not automatically reset to 0. Default: FULL-HOUR | |||||||||||||||||||||
SYSLOG-SIZE= | size Automatic size monitoring of the system log file SYSLOG by openUTM.
The administrator can modify the generated threshold value, and activate or deactivate size monitoring as desired during operation (e.g. with the KDCSLOG command). Default: 0 (no size monitoring) | |||||||||||||||||||||
SYSTEM-TASKS | Controls the number of UTM system processes. To ensure that an application continues to be responsive and, for example, can also process internal jobs to terminate transactions, communication between the nodes of a UTM cluster application or an administrator's jobs in these situations, UTM starts further processes, known as UTM system processes, in addition to the processes generated and started by the user. Generally no program unit runs are executed by the UTM system processes, and they are only used for internal jobs in bottleneck situations. Consequently the additional UTM system processes only place a slight load on the host. | |||||||||||||||||||||
*STD | *STD means that UTM starts (up to) three additional processes for the application; these are then used as UTM system processes. Depending on the number of tasks started for the application, the second, fourth, and seventh processes of an application become UTM system processes. *STD is the default value. | |||||||||||||||||||||
number | Maximum number of UTM system processes which are to be started in addition for the application. The value 0 means that no UTM system process will be started. Minimum value: 0 Values greater than 10 are ignored and rounded down to 10. The table below shows how many UTM system processes are started additionally for the generated value SYSTEM-TASKS=*STD in accordance with the start parameter TASKS:
If more than three UTM system processes are generated, depending on the value of SYSTEM-TASKS and the number of started processes, the 11th, 21st, 31st, 41st, 51st, 61st, and 71st processes also become UTM system processes. | |||||||||||||||||||||
TASKS= | number Maximum number of processes that can be used simultaneously for the application. This is a mandatory operand. Minimum value: 2 KDCDEF automatically resets values < 2 to 2 without outputting a UTM message. The current number of processes is defined when starting the application. You can specify TASKS=1 during startup. The administrator can dynamically modify the number of processes during runtime (e.g. with the administration command KDCAPPL). The number of processes specified during startup or set by the administrator must not exceed the value generated here. | |||||||||||||||||||||
TASKS-IN-PGWT= | number Maximum number of processes of the UTM application in which program units with blocking calls, e.g. the KDCS call PGWT, may run simultaneously. The value of TASKS-IN-PGWT must be less than that of the TASKS= operand. If TASKS-IN-PGWT=0, it is not possible to generate a TAC class or a transaction code (TAC) for which blocking calls are permitted (see TAC/TACCLASS ...,PGWT=). In this case, PGWT=NO must be specified in all TACCLASS and TAC statement (see also the TAC statement in section"TAC - define the properties of transaction codes and TAC queues" and the TACCLASS statement in section "TACCLASS - define the number of processes for a TAC class" for more information). Default value: 0 | |||||||||||||||||||||
TERMWAIT= | time (terminal wait) Default: 600 | |||||||||||||||||||||
TRACEREC= | number Maximum number of entries in the process-specific trace areas handled by openUTM. This value applies to the trace area
openUTM writes trace information to these areas for diagnostic purposes. Length of the entries:
Default: 32500 KDCDEF automatically resets values < 1 to the default value and values > 32500 to the maximum value without outputting a UTM message. | |||||||||||||||||||||
TRMSGLTH= | length This defines the maximum value for the following:
Default: 32700 bytes Maximum value: 32700 bytes If you use RSO printers, the size of the RSO buffer (REMOTE-BUFFER- SIZE in the SPOOL parameter file) must be greater than or equal to 32700. See also section "Defining the RSO buffer size" (Entries for RSO and SPOOL) for more information. | |||||||||||||||||||||
USLOG= | This defines single- or dual-file operation for the user log file USLOG. | |||||||||||||||||||||
SINGLE | Single-file operation is activated for the user log file. Default: SINGLE | |||||||||||||||||||||
DOUBLE | For security reasons, dual-file operation is activated for the user log file. | |||||||||||||||||||||
VGMSIZE= | number This operand is only supported on BS2000 systems. This parameter is used to generate a buffer area with the specified size for the service memory of an SQL database system. It also restricts the user’s share of the page pool. VGMSIZE= is specified in KB. If the service memory area to be logged when the PEND call is issued is greater than number, the service is terminated with PEND ER. Default value: 32KB | |||||||||||||||||||||
XAPTPSHMKEY= | number This operand is only supported on Unix, Linux and Windows systems. Authorization key for the XAPTP shared memory segment Shared memory keys are global system parameters. XAPTPSHMKEY is a mandatory operand if the application is to communicate via the OSI TP protocol. | |||||||||||||||||||||
The table below provides an overview of the purpose and default values of the individual operands of the MAX statement:
Operand | Purpose | Mandatory | Default value |
|---|---|---|---|
Operands valid for all operating systems | |||
APPLIMODE= | Choice of UTM variant: UTM-S or UTM-F | SECURE | |
APPLINAME= | Name of the UTM application | X | - |
ASYNTASKS= | Asynchronous processing (number of processes for asynchronous processing and asynchronous services open at the same time) | 1, 1 | |
BLKSIZE= | Size of a UTM page | 2K (in UTM cluster | |
CACHESIZE= | Tuning feature (size and properties of the cache) | Depending on the system: | |
CLRCH= | Character for overwriting the communication area and standard primary working area | None | |
CONN-USERS= | Restriction on the number of users or clients active simultaneously | Depending on the system: | |
CONRTIME= | Automatic connection setup for printers (waiting time for reconnection) | 10 minutes | |
DATA-COMPRESSION | Controlling data compression | STD | |
DEAD-LETTER-Q-ALARM= | Monitors the number of messages received in the dead letter queue | 0, i.e. no monitoring | |
DESTADM= | Asynchronous administration | None | |
DPUTLIMIT1= | Time-driven jobs (upper limit) | 360 days | |
DPUTLIMIT2= | Time-driven jobs (lower limit) | 1 day | |
GSSBS= | GSSB storage areas (maximum number) | 32 | |
HOSTNAME= | Virtual host name for the UTM application | 8 blanks | |
KB= | Maximum length of the communication area | 512 | |
KDCFILE= | Assigning a KDCFILE | X | - |
KEYVALUE= | Data access control using the lock/key code concept (number of the highest key code) | 32 | |
LEADING-SPACES= | Pass leading blanks in messages from terminals or from TS applications | NO | |
LPUTBUF= | Logging of user data with LPUT (number of PAM pages in the page pool) | 1 | |
LPUTLTH= | Logging of user data with LPUT (maximum LPUT message length) | 1948 bytes | |
LSSBS= | LSSB storage areas (maximum number) | 8 | |
MOVE-BUNDLE-MSGS=NO | Automatic moving of waiting asynchronous messages of a slave LTERM, a slave LPAP, or a slave OSI-LPAP | NO | |
NB= | Maximum length of the KDCS message area | 2048 | |
NRCONV= | Maximum number of stacked services | 0 | |
OSI-SCRATCH-AREA= | Size in KB of an internal UTM working area | 256 | |
PGPOOL= | Size of the page pool and warning levels | 100 UTM pages, 80%, 95% | |
PGPOOLFS= | Tuning feature: splitting the page pool | Page pool in KDCFILE | |
PGWTTIME= | Maximum time for the KDCS call PGWT | TERMWAIT=time | |
PRIVILEGED-LTERM= | Define the privileged LTERM | - | |
QTIME | Maximum permitted wait time for messages from service controlled queues | 32767 seconds | |
RECBUF= | Tuning feature: | 5 PAM pages per process, | |
RECBUFFS= | Tuning feature: | in KDCFILE | |
REDELIVERY= | Maximum number of redeliveries of an asynchronous message | 0 for UTM-controlled queues, | |
RESWAIT= | Waiting time for a resource (e.g. GSSB, TLS) locked by another transaction (time1) or process (time2) | 120 seconds, 300 seconds | |
SPAB= | Maximum SPAB length | 512 | |
SM2= | Permitting, activating, and deactivating the supply of UTM data to SM2 | OFF | |
STATISTICS-MSG= | Statistics message K081 is produced and the counter is automatically reset to 0 | FULL-HOUR | |
SYSLOG-SIZE= | Automatic size monitoring of the SYSLOG file by openUTM | 0 | |
SYSTEM-TASKS | Number of UTM system processes | *STD | |
TASKS= | Number of UTM processes | X | - |
TASKS-IN-PGWT= | Number of processes for PGWT jobs | 0 | |
TERMWAIT= | Maximum waiting time for dialog input within a transaction | 600 seconds | |
TRACEREC= | Space reserved for diagnostic information (number of entries) | 32500 | |
TRMSGLTH= | Maximum message length | 32700 bytes | |
USLOG= | Single- or dual-file operation of the user log file | SINGLE | |
VGMSIZE= | Generate the buffer area with the specified size | 32 KB | |
| BS2000-specific operands | |||
BRETRYNR= | Communication with BCAM (number of retries when sending messages) | 10 | |
CARDLTH= | ID card reader for KDCSIGN check | 0 | |
CATID= | Catalog IDs for the KDCFILE | Default CATID | |
LOCALE= | Default language environment | Blanks | |
LOGACKWAIT= | Support for output devices (waiting time for confirmation) | 600 seconds | |
MP-WAIT= | Maximum waiting time per process for connection to the common memory pool | 180 seconds | |
PRINCIPAL-LTH= | Maximum length of a Kerberos principal in bytes | 32 | |
REQNR= | Tuning feature: PAM I/O jobs (maximum number of parallel jobs) | 20 | |
SAT= | Minimum logging of events with SAT | OFF | |
VGMSIZE= | Size of the buffer area for the service memory of an SQL database system | 32KB | |
| Unix, Linux and Windows system-specific operands | |||
CACHESHMKEY= | Authorization key for a shared memory segment (global buffer for file access) | X | - |
IPCSHMKEY= | Authorization key for a shared memory segment (communication between UTM processes) | X | - |
IPCTRACE= | Number of UTM entries in the IPC trace area | 1060 | |
KAASHMKEY= | Authorization key for a shared memory segment (global data) | X | - |
OSISHMKEY= | Authorization key for an OSS shared memory segment | with | - |
SEMARRAY= | Range of semaphore keys for global semaphores (alternative to SEMKEY) | X | - |
SEMKEY= | Semaphore keys for global semaphores (alternative to SEMARRAY) | X | - |
XAPTPSHMKEY= | Authorization key for the XAPTP shared memory segment | with | - |