Tuning and logging Beyond the administration activities described in other chapters, there are other things you can do to tune or troubleshoot your system. Child Topics:Virtual memory modification If many simultaneous users intend to perform heavy data pivoting via Information Services or in other ways stress the server, you may need to modify the amount of memory available to the virtual computer. Garbage collection loggingWhen old requests to a service become obsolete, the objects created in memory becomes garbage. By enabling garbage collection logs (GC logs) you can get an understanding of your system's performance and troubleshoot memory issues. Custom configurations for managing space needs If you need more space for library content, log files, information links, or the files that the Web Player service writes to the hard disk, you can change the default settings to store these items in different directories. Temporary tablespace By default, the tablespaces/database files for Spotfire Server with either an Oracle or SQL database uses autoextend/autogrowth. If this does not meet your needs, alter the settings. Increasing the number of available sockets on LinuxSpotfire Server opens many connections, and each requires a file descriptor. For performance and security reasons, Linux has limited the number of connections that can be opened by a process. You may want to increase this limit. Changing the settings that determine when Web Player instances are recycled due to low temporary disk space By default, Spotfire recycles Web Player instances if their temporary disk space falls below 1500 MB (the "exhausted" level), and remains below that level for one hour. When a service instance is recycled, the corresponding process is restarted; all open analyses on that instance are closed and its temp files are removed. You can change the definition of "exhausted" and the period of time that trigger Web Player recycling. Restarting a node manager to terminate its running jobs Use this procedure to "refresh" a node when its service instances appear to be running jobs that should have terminated. Virtual memory modification If many simultaneous users intend to perform heavy data pivoting via Information Services or in other ways stress the server, you may need to modify the amount of memory available to the virtual computer. Garbage collection loggingWhen old requests to a service become obsolete, the objects created in memory becomes garbage. By enabling garbage collection logs (GC logs) you can get an understanding of your system's performance and troubleshoot memory issues. Custom configurations for managing space needs If you need more space for library content, log files, information links, or the files that the Web Player service writes to the hard disk, you can change the default settings to store these items in different directories. Temporary tablespace By default, the tablespaces/database files for Spotfire Server with either an Oracle or SQL database uses autoextend/autogrowth. If this does not meet your needs, alter the settings. Increasing the number of available sockets on LinuxSpotfire Server opens many connections, and each requires a file descriptor. For performance and security reasons, Linux has limited the number of connections that can be opened by a process. You may want to increase this limit. Changing the settings that determine when Web Player instances are recycled due to low temporary disk space By default, Spotfire recycles Web Player instances if their temporary disk space falls below 1500 MB (the "exhausted" level), and remains below that level for one hour. When a service instance is recycled, the corresponding process is restarted; all open analyses on that instance are closed and its temp files are removed. You can change the definition of "exhausted" and the period of time that trigger Web Player recycling. Restarting a node manager to terminate its running jobs Use this procedure to "refresh" a node when its service instances appear to be running jobs that should have terminated. Parent topic: Advanced proceduresHomeAdvanced proceduresTuning and logging
Tuning and logging Beyond the administration activities described in other chapters, there are other things you can do to tune or troubleshoot your system. Child Topics:Virtual memory modification If many simultaneous users intend to perform heavy data pivoting via Information Services or in other ways stress the server, you may need to modify the amount of memory available to the virtual computer. Garbage collection loggingWhen old requests to a service become obsolete, the objects created in memory becomes garbage. By enabling garbage collection logs (GC logs) you can get an understanding of your system's performance and troubleshoot memory issues. Custom configurations for managing space needs If you need more space for library content, log files, information links, or the files that the Web Player service writes to the hard disk, you can change the default settings to store these items in different directories. Temporary tablespace By default, the tablespaces/database files for Spotfire Server with either an Oracle or SQL database uses autoextend/autogrowth. If this does not meet your needs, alter the settings. Increasing the number of available sockets on LinuxSpotfire Server opens many connections, and each requires a file descriptor. For performance and security reasons, Linux has limited the number of connections that can be opened by a process. You may want to increase this limit. Changing the settings that determine when Web Player instances are recycled due to low temporary disk space By default, Spotfire recycles Web Player instances if their temporary disk space falls below 1500 MB (the "exhausted" level), and remains below that level for one hour. When a service instance is recycled, the corresponding process is restarted; all open analyses on that instance are closed and its temp files are removed. You can change the definition of "exhausted" and the period of time that trigger Web Player recycling. Restarting a node manager to terminate its running jobs Use this procedure to "refresh" a node when its service instances appear to be running jobs that should have terminated. Virtual memory modification If many simultaneous users intend to perform heavy data pivoting via Information Services or in other ways stress the server, you may need to modify the amount of memory available to the virtual computer. Garbage collection loggingWhen old requests to a service become obsolete, the objects created in memory becomes garbage. By enabling garbage collection logs (GC logs) you can get an understanding of your system's performance and troubleshoot memory issues. Custom configurations for managing space needs If you need more space for library content, log files, information links, or the files that the Web Player service writes to the hard disk, you can change the default settings to store these items in different directories. Temporary tablespace By default, the tablespaces/database files for Spotfire Server with either an Oracle or SQL database uses autoextend/autogrowth. If this does not meet your needs, alter the settings. Increasing the number of available sockets on LinuxSpotfire Server opens many connections, and each requires a file descriptor. For performance and security reasons, Linux has limited the number of connections that can be opened by a process. You may want to increase this limit. Changing the settings that determine when Web Player instances are recycled due to low temporary disk space By default, Spotfire recycles Web Player instances if their temporary disk space falls below 1500 MB (the "exhausted" level), and remains below that level for one hour. When a service instance is recycled, the corresponding process is restarted; all open analyses on that instance are closed and its temp files are removed. You can change the definition of "exhausted" and the period of time that trigger Web Player recycling. Restarting a node manager to terminate its running jobs Use this procedure to "refresh" a node when its service instances appear to be running jobs that should have terminated. Parent topic: Advanced proceduresHomeAdvanced proceduresTuning and logging