Summary: The ALL-In-Cache Cache policy
OverviewWhen running in ALL_IN_CACHE cache policy mode, the space uses only the available physical memory. When running in a persistent space mode and having Space Persistency defined, the space data is backed with the underlying database, but the overall capacity of the space does not exceed the capacity of the available physical memory.
SpaceMemoryShortageExceptionThe org.openspaces.core.SpaceMemoryShortageException (which wraps the com.j_spaces.core.MemoryShortageException) is thrown when:
Here is an example of the org.openspaces.core.SpaceMemoryShortageException message: org.openspaces.core.SpaceMemoryShortageException at: host: MachineHostName, container: mySpace_container1_1, space mySpace, total memory: 1820 mb, used memory: 1283 mb Getting SpaceMemoryShortageException when running in ALL-IN_CACHE policy mode points that the heap size allocated for the JVM is too low, or the garbage collection settings are not tuned correctly. You should consider increasing the JVM heap size or move into more advanced garbage collector such as CMS or iCMS. Here is an example for CMS garbage collector settings you may use: -server -Xms8g -Xmx8g -Xmn300m -XX:+UseConcMarkSweepGC -XX:+UseParNewGC -XX:CMSInitiatingOccupancyFraction=60 -XX:+UseCMSInitiatingOccupancyOnly -XX:MaxPermSize=256m -XX:+ExplicitGCInvokesConcurrent -XX:+UseCompressedOops When running with large heap size you should tune the memory manager settings. This will avoid throwing SpaceMemoryShortageException too early. See the Handling Large JVM Heap Size section for details. |
![]() |
GigaSpaces.com - Legal Notice - 3rd Party Licenses - Site Map - API Docs - Forum - Downloads - Blog - White Papers - Contact Tech Writing - Gen. by Atlassian Confluence |