Administration guide > Plan the WebSphere eXtreme Scale environment > Cache topology
WebSphere eXtreme Scale grids can run within existing processes as embedded eXtreme Scale servers or you can manage them as external processes.
Embedded grids are useful when you are running in an application server, such as WAS. You can start eXtreme Scale servers that are not embedded by using command line scripts and run in a Java™ process.
Figure 1. Embedded cache
- Simplified administration since there are less processes to manage.
- Simplified application deployment since the grid is using the client application classloader.
- Supports partitioning and high availability.
- Increased the memory footprint in client process since all of the data is collocated in the process.
- Increase CPU utilization for servicing client requests.
- More difficult to handle application upgrades since clients are using the same application Java archive files as the servers.
- Less flexible. Scaling of clients and grid servers cannot increase at the same rate. When servers are externally defined, you can have more flexibility in managing the number of processes.
When to use
Use embedded grids when there is plenty of memory free in the client process for grid data and potential failover data.
For more information, see Enable the client invalidation mechanism.
Parent topic:Cache topology: In-memory and distributed caching
Local in-memory cache
Peer-replicated local cache
Multi-master data grid replication topologies