(Chapter 23, section 3)
Even better than rapidly resolving a problematic situation is preventing the situation from occurring at all. There are several recurring tasks to perform as part of the operations around the SOA Suite to preempt problems. These are typically focused on optimal use of system resources. This includes balancing the load over the available resources and preventing unmanageable loads from even reaching the SOA infrastructure. It also entails freeing up resources as quickly as possible – for example by timely removal of no longer needed remnants of past activity.
In this section, we look at some of the things to do and use given a certain infrastructure layout and capacity: how to protect the infrastructure from total collapse under a peak load, how to influence fair sharing of resources and how to clean up after ourselves to ensure continued smooth operations. Throttling, [system] resource allocation and purging are the main topics in this section.
This figure shows some the topics discussed in this section.
The sources for this section are found in folder ch23.
A-Team blog: Throttling in SOA Suite via Parking Lot Pattern
Blog by AVIO – Throttling in API Gateway
Albin’s Blog: Message Throttling using Database Adapter
Anthony Reynold’s Following the thread in OSB
Vladimir Dyuzhev: Work Manager Is Not A Thread Pool; Max Threads Constraint Is
Tjeerd Kaastra: Restricting the number of JMS connections made by the OSB