With the development of Sky-Walking 2.0-2016 version, there are a large number of memory-level, non-strict FIFO publish/subscribe queue requirements in each module. Therefore, we created a new project DataCarrier: a lightweight, embedded, high-speed processing publish-subscribe queue library.
The project is based on jdk1.6 and provides wider code support.
DataCarrier is very lightweight, with less than 20k releases, and does not depend on any third-party components.
Adopt multi-channel, fixed memory allocation mode. DataCarrier will not increase the memory capacity during operation, support multi-threaded concurrent data generation, and provide multi-threaded consumption batch consumption capability.
Multiple DataCarrier instances can be declared at runtime for multiple different scenarios.
DataCarrier is derived from the core model of SkyWalking, after a long-running test, version 1.0 will also be directly used by Sky-Walking version 2.0-2016.
The homepage of DataCarrier provides simple and clear usage examples, as well as a large number of TestCases, which are convenient for everyone to start and use quickly.
Project homepage and documentation page: https://github.com/wu-sheng/DataCarrier