Product overview > Scalability overview

Scaling in units or pods

Although you can deploy a data grid over thousands of JVMs, you might consider splitting the data grid into units or pods to increase the reliability and ease of testing of the configuration. A pod is a group of servers that is running the same set of applications.

Deploy a large single data grid

Test has verified that eXtreme Scale can scale out to over 1000 JVMs. Such testing encourages building applications to deploy single data grids on large numbers of boxes. Although it is possible to do this, it is not recommended, for several reasons:

  1. Budget concerns: Your environment cannot realistically test a 1000-server data grid. However, it can test a much smaller data grid considering budget reasons, so you do not need to buy twice the hardware, especially for such a large number of servers.

  2. Different application versions: Requiring a large number of boxes for each testing thread is not practical. The risk is that you are not testing the same factors as you would in a production environment.

  3. Data loss: Running a database on a single hard drive is unreliable. Any problem with the hard drive causes you to lose data. Running a growing application on a single data grid is similar. You will likely have bugs in the environment and in the applications. So placing all of the data on a single large system will often lead to a loss of large amounts of data.

Splitting the data grid

Splitting the application data grid into pods (units) is a more reliable option. A pod is a group of servers that are running a homogenous application stack. Pods can be of any size, but ideally they should consist of about 20 physical servers. Instead of having 500 physical servers in a single data grid, you can have 25 pods of 20 physical servers. A single version of an application stack should run on a given pod, but different pods can have their own versions of an application stack.

Generally, an application stack considers levels of the following components.

A pod is a conveniently sized deployment unit for testing. Instead of having hundreds of servers for testing, it is more practical to have 20 servers. In this case, you are still testing the same configuration as you would have in production. Production uses grids with a maximum size of 20 servers, constituting a pod. You can stress-test a single pod and determine its capacity, number of users, amount of data, and transaction throughput. This makes planning easier and follows the standard of having predictable scaling at predictable cost.

Set up a pod-based environment

In different cases, the pod does not necessarily have to have 20 servers. The purpose of the pod size is for practical testing. The size of a pod should be small enough that if a pod encounters problems in production, the fraction of transactions affected is tolerable.

Ideally, any bug impacts a single pod. A bug would only have an impact on four percent of the application transactions rather than 100 percent. In addition, upgrades are easier because they can be rolled out one pod at a time. As a result, if an upgrade to a pod creates problems, the user can switch that pod back to the prior level. Upgrades include any changes to the application, the application stack, or system updates. As much as possible, upgrades should only change a single element of the stack at a time to make problem diagnosis more precise.

To implement an environment with pods, you need a routing layer above the pods that is forwards and backwards compatible if pods get software upgrades. Also, you should create a directory that includes information about which pod has what data. Use another eXtreme Scale data grid for this with a database behind it, preferably using the write-behind scenario.) This yields a two-tier solution. Tier 1 is the directory and is used to locate which pod handles a specific transaction. Tier 2 is composed of the pods themselves. When tier 1 identifies a pod, the setup routes each transaction to the correct server in the pod, which is usually the server holding the partition for the data used by the transaction. Optionally, you can also use a near cache on tier 1 to lower the impact associated with looking up the correct pod.

Use pods is slightly more complex than having a single data grid, but the operational, testing, and reliability improvements make it a crucial part of scalability testing.

Parent topic:

Scalability overview

Related concepts

Data grids, partitions, and shards


Placement and partitions

Single-partition and cross-data-grid transactions

Related tasks

Upgrade and migrating WebSphere eXtreme Scale