SAP profile parameters

Edit online

The tables in this information unit list and describe the profile parameters that are related to ASCS/SCS.

Table 1. Enqueue client parameters relevant for the high availability solution
Parameter Description Default value Recommended value
ms/persist_lg_info Information on logon groups can be stored regularly and during stop operations of the message server. It can be restored after its restart so that users and RFC connections by using these logon groups can directly log on after the message server restart.

Kernel 7.2x:
OFF
Kernel 7.4x:
ON

SAP Kernel 7.2x: ON
SAP Kernel 7.4x: no
need to set it, since
the default value is: ON

This parameter setting assumes that the recommended file system setup is done on the SAP Central Services instance directories on a shared zFS file system (see Setting up zFS file systems). The message server creates the file $(DIR_HOME)/<SID>_msg_server_adtl_storage and updates it on a regular basis. For more details, see SAP Note 1787163: Message Server: save logon groups feature.
enque/serverhost Host of the enqueue server.   <virtual hostname>
enque/serverinst Instance number of the enqueue server.   <instance number>
enque/process_location Specifies where the enqueue requests are processed. OPTIMIZE REMOTESA (for application servers)
enque/deque_wait_answer Indicates whether a dequeue request waits for the acknowledgment. If the default value (FALSE) is used, obsolete locks might remain in the enqueue table on failover and must be removed manually. If TRUE is specified, the reported enqueue time of all transactions increases slightly. FALSE TRUE
enque/sync_dequeall When an SAP transaction is ended, the system uses a final DequeueAll() to clean up the system resources. This DequeueAll() is an asynchronous request, where a chance remains that the system assumes its DequeueAll() was sent, although it may never arrive at the stand-alone enqueue server, for example, if a network problem occurs. The related locks are retained in such a case, which does not result in inconsistencies, but could block other transactions. 0 1
enque/server/repl/async Indicates whether replication of enqueue locks should be done in an asynchronous mode. If this parameter is set to TRUE, this may cause a loss of enqueue locks in case the enqueue server fails before the locks were sent - asynchronously - to the replication server. (see SAP Note 2036171: Missing locks following failover between enqueue server and replicated enqueue server ).

For high availability systems, the parameters must be set to FALSE.

Kernel 7.20:
TRUE
Kernel 7.21
and later:
FALSE
Kernel 7.4x:
FALSE

SAP kernel 7.20:
explicitly set to FALSE
SAP kernel 7.21 or higher:
no need to explicitly set
the parameter, since the
default value is OK.
enque/con_timeout Specifies the maximum time (in milliseconds) the enqueue client can block, when a connection is being set up. If the connection cannot be established, the client waits for one second and then retries. Parameter enque/con_retries (see next row) determines how often the attempts are repeated. 5000 5000

In order to get a faster reconnection in case of a planned failover of the enqueue server, the parameter value should be 1000. This means, the client tries each two seconds to reconnect according to the formula described for the enque/con_retries parameter.
enque/con_retries Specifies the maximum number of times the enqueue client tries to establish a connection to the enqueue server.
After the specified number of retries, the caller (the application) reports an error. The following formulas calculate the interval of the entire wait time: 
1(sec) * <enque/con_retries>
is the lower value of the interval and 

1(sec) + (<enque/con_timeout> / 1000) 
      * <enque/con_retries>
is the upper value.

If you keep the default values, this means that an error is triggered after 301 seconds.

Set this parameter to a value so that the calculated result from the mentioned parameters is larger than the typical time for an ASCS or SCS instance failover. This depends largely on your operating system and your cluster solution. If a failover can take longer, you should increase the value of these parameters.

With the settings recommended in this document, the time is 8 minutes and 21 seconds (501 seconds).

 60 100
Table 2. Enqueue server parameters relevant for the high availability solution
Parameter Description Default value Recommended value
enque/server/replication Enables replication. false true
enque/encni/repl_port Port number, which the enqueue server opens for replication.

Use this parameter if the replication port of the enqueue server cannot be set to the default port 5XX16 (where XX is the instance number). It should then be set either in DEFAULT.PFL, or in the profiles of the enqueue server and the replication server.

 default port 5XX16 where XX is the instance number. 5XX16

If possible, do not specify this parameter. The default port 5XX16 is used in this case.

enque/server/threadcount Number of I/O threads in the enqueue server. 1 1
enque/backup_file Specifies the complete path to the enqueue backup file. This is used to save the locks, which are transferred to the update, in case the enqueue server is intentionally stopped (including a stop of the replication server). If the enqueue server is restarted it reads the locks from this file again, so that updates, which have not been processed can be processed. If a shared file system is used, such as zFS under z/OS® UNIX, the default value is satisfactory.
Note: If ASCS and/or SCS runs under Linux® or AIX® and no shared file system is used for the instance directory, it must be set to: 
enque/backup_file = $(DIR_GLOBAL)/ENQBCK
 /usr/sap/<SID>/ASCS<instance number>/log/ENQBCK

or

/usr/sap/<SID>/SCS<instance number>/log/ENQBCK