Bookshelf Home | Contents | Index | Search | PDF | ![]() ![]() ![]() ![]() |
Siebel Analytics Installation and Configuration Guide > NQSConfig.ini File Reference > Parameter and Syntax Descriptions >
Parameters in the Server Section
The parameters in the Server section define defaults and limits for the Siebel Analytics Server.
SERVER_NAME
A logical name identifying the Siebel Analytics Server.
Example:
SERVER_NAME = Siebel_Analytics_Server ;
CLIENT_SERVER_SHARED_MEMORY_SIZE
The size of the shared memory block dedicated to each request and session connecting to the server.
Example:
CLIENT_SERVER_SHARED_MEMORY_SIZE = 32 KB ;
SESSION_TIME_OUT_SECONDS
Specifies the time, in seconds, that a session will wait for a connection to occur before terminating.
Example:
SESSION_TIME_OUT_SECONDS = 300 ;
MAX_SESSION_LIMIT
Specifies the maximum number of connections allowed by the server. If this number is exceeded, the server will refuse the connection request. The limit is 65,535 connections.
Example:
MAX_SESSION_LIMIT = 100 ;
MAX_REQUEST_PER_SESSION_LIMIT
Specifies the maximum number of logical requests per session. This is how many open requests there will be, per session, at the same time. The limit is 65535 logical requests per session.
NOTE: Usually, individual users will have only one open request per session at the same time. Application programs and Siebel Analytics Web, however, will typically have more than one open at the same time. In general, the default value of 100 should be sufficient for most environments, but this parameter should be tuned based on the application environment and the client tools in use.
Example:
MAX_REQUEST_PER_SESSION_LIMIT = 100 ;
SERVER_THREAD_RANGE
For each Siebel Analytics Server request, this specifies configuration information for thread allocation. The lower bound of the range specifies the number of threads initially allocated, and the upper bound of the range specifies the maximum number of threads to be allocated. The thread pool grows and shrinks in 5 thread increments until the upper or lower bound is reached. If there are fewer threads than sessions, sessions will share the available number of threads on a first come-first served basis.
Set both values the same to maximize the benefits of thread pooling. The default value (40-40) eliminates thread start and exit activities, which can degrade performance.
Example:
SERVER_THREAD_RANGE = 40-40 ;
SERVER_THREAD_STACK_SIZE
Specifies the memory stack size allocated for each server thread. The default of 0 sets the stack size as 1 MB in Windows NT.
Example:
SERVER_THREAD_STACK_SIZE = 0 MB ;
DB_GATEWAY_THREAD_STACK_SIZE
Specifies the memory stack size allocated for each the Siebel Analytics Server COM Gateway thread. The default of 0 sets the stack size as 1 MB in Windows NT.
Example:
DB_GATEWAY_THREAD_STACK_SIZE = 0 KB ;
MAX_EXPANDED_SUBQUERY_PREDICATES
This controls the maximum number of values that may be in an IN value list populated by a subquery. The default is 8,192 values.
Example:
MAX_EXPANDED_SUBQUERY_PREDICATES = 8192;
MAX_QUERY_PLAN_CACHE_ENTRIES
Controls the number of cached compiler query plans. The query plan cache is an internal performance feature that increases the speed of the query compilation process by caching plans for the most recently used queries.
Example:
MAX_QUERY_PLAN_CACHE_ENTRIES = 10240 ;
MAX_DRILLDOWN_INFO_CACHE_ENTRIES
Controls the number of cached drill-down information entries per repository. This increases the speed of computing drill-down information by caching the drill-down information for the most recently used queries.
Example:
MAX_DRILLDOWN_INFO_CACHE_ENTRIES = 10240 ;
MAX_DRILLDOWN_QUERY_CACHE_ENTRIES
Controls the number of cached drill-down query entries per repository. This increases the speed of drilling down by caching the drill-down results for the most recently used queries.
Example:
MAX_DRILLDOWN_QUERY_CACHE_ENTRIES = 10240 ;
CLUSTER_PARTICIPANT
Specifies whether the Siebel Analytics Server that is using this configuration file is a member of a Siebel Analytics Server cluster. Valid values are YES and NO. The default value is NO.
If the server is to be a member of a Siebel Analytics Server cluster, uncomment the parameters REPOSITORY_PUBLISHING_DIRECTORY and REQUIRE_PUBLISHING_DIRECTORY, and supply valid values for them.
NOTE: When CLUSTER_PARTICIPANT is set to YES, this server needs to have a valid, configured NQClusterConfig.ini file in the Config subdirectory, located in the Siebel Analytics software installation directory. For more information, see NQClusterConfig.ini File Reference.
Example:
CLUSTER_PARTICIPANT = YES;
REPOSITORY_PUBLISHING_DIRECTORY
When the parameter CLUSTER_PARTICIPANT is set to YES, this specifies the location of the repository publishing directory shared by all Siebel Analytics Servers participating in the cluster. There is no default value for this parameter.
When a repository is updated in online mode, it is published to this location. All clustered servers will examine this location upon startup for any repository changes. This needs to be a valid location visible to all servers in the cluster even if you anticipate that no repositories will be updated in online mode.
The directory should reside on a shared file system. The directory needs to be a valid fully-qualified directory pathname, with double quotes ( " ) surrounding the pathname. Both UNC and mapped directories are supported. Do not specify a relative path name, or the Repository subdirectory (located in the Siebel Analytics software installation directory) as the location of the repository publishing directory.
This entry is required on every Siebel Analytics Server that is to participate in the cluster. The Siebel Analytics Server designated as the master server for online repository changes (from the parameter MASTER_SERVER in the NQClusterConfig.ini file) needs to have read and write access to this directory. The Siebel Analytics Servers in the cluster (from the parameter SERVERS in the NQClusterConfig.ini file) need to have read and write access to this directory as well. All entries need to reference the same actual directory, although different names can be specified to accommodate differences in drive mappings.
These are examples:
REPOSITORY_PUBLISHING_DIRECTORY = "z:\SiebelAnalytics\Publish";
REPOSITORY_PUBLISHING_DIRECTORY = "\\ClusterSrv\Publish";REQUIRE_PUBLISHING_DIRECTORY
When the parameter CLUSTER_PARTICIPANT is set to YES, this specifies that the repository publishing directory (from the parameter REPOSITORY_PUBLISHING_DIRECTORY) needs to be available in order for this Siebel Analytics Server to start up and join the cluster. The default value is YES. When set to YES, if the publishing directory is not available at startup or if an error is encountered while the server is reading any of the files in the directory, an error message will be logged in the NQServer.log file and the server will shut down.
To allow this Siebel Analytics Server to start up and join the cluster even if the publishing directory is not available, set this value to NO. When set to NO, the server will join the cluster and a warning message will be logged in the NQServer.log file. Any online repository updates will not be reflected in the server's Repository directory (located in the Siebel Analytics software installation directory). This could result in request failures, wrong answers, and other problems. However, this could be useful in situations where online repository editing will be done infrequently and the goal is to keep the cluster operational even if some servers have stale repositories.
Example:
REQUIRE_PUBLISHING_DIRECTORY = YES;
INIT_BLOCK_CACHE_ENTRIES
Controls the number of initialization block result sets that are cached with respect to row-wise initialization. The cache key is the fully instantiated initialization block SQL. The default value is 20. Because this parameter affects internal operations for localized versions of the Siebel Analytics software, we recommend that you do not change this value unless instructed to do so.
Example:
INIT_BLOCK_CACHE_ENTRIES = 20 ;
CLIENT_MGMT_THREADS_MAX
This specifies the number of management threads to allocate for managing Siebel Analytics Server client/server communications. The default value of 5 is sufficient for server communications with the Administration Tool and when Siebel Analytics Web is the only main client process.
Each client process consumes a management thread. If you plan to run multiple Analytics Web client instances or multiple third-party client instances, increase the number of management threads to match the number of client instances. If the number of running client instances exceeds the number of management threads allocated, worker threads will be consumed.
Example:
CLIENT_MGMT_THREADS_MAX= 10 ;
RPC_SERVICE_OR_PORT
This specifies the Remote Procedure Call (RPC) Service or TCP/IP port the Siebel Analytics Server should use for client/server communications. The default is 9703.
When using the Analytics ODBC Wizard to set up ODBC data sources for the Siebel Analytics Server, the port number specified in the Port field on the Enter Logon Information screen should match the port number specified here. If you change the port number in the configuration file, remember to reconfigure any affected ODBC data sources to use the new port number.
Example:
RPC_SERVICE_OR_PORT= 9703 ;
ENABLE_DB_HINTS
This enables optional hints to be passed along with a SQL statement to an Oracle database. Database hints are discussed in Siebel Analytics Server Administration Guide.
PREVENT_DIVIDE_BY_ZERO
This parameter controls the behavior for when a division by zero occurs. If set to Yes, then a NULL value is returned. If set to No, then the query is terminated and an appropriate error is returned to the user.
Bookshelf Home | Contents | Index | Search | PDF | ![]() ![]() ![]() ![]() |
Siebel Analytics Installation and Configuration Guide, Version 7.5, Rev. C Published: 18 April 2003 |