We started Java Cá&Lá (JCL for short) in 2010, after many years using several middlewares for High Performance Computing (HPC), like RMI, Corba, Map reduce, JPFF, Ice and several others. We started a new middleware solution for HPC, because the existing solutions were and still are very complex for programming, time consuming for deployment, usually require huge amount of refactorings and put together the business logic with the distribution issues. Java was selected, since it is, together with Android, a huge development community. In 2010, JCL was born with HPC goals (processing and storage services). JCL supports the distributed shared memory (DSM) and task-oriented programming models, regardaless we are using a massive parallel server or a cluster of commodity PCs.


In 2011, the first undergradute student graduated using JCL as his investigation theme. 


JCL was used by four other undergraduate students as they investigation themes. Research projects were conducted to make JCL a collaborative middleware solution. The first small and simple JCL API was finally achieved. The deployment process became very simple. Some general purpose JCL applications were developed. 


The first master student accepted in JCL. The thesis goal was to produce the first middleware solution that puts together several features of HPC solutions presented separately in the last decades. Super-peer, task scheduling, a native distributed hash map, a simple deploy, an API to analyze a task costs during its execution over a cluster, a fair storage allocation mechanism and several other features were implemented and tested exhaustively. Many JCL applications were coded to test the API utility and simplicity.


The first master student finished his work and the first JCL paper was published. Our team became bigger with five undergraduate students helping the tests and starting new research investigations. The second master student joined the JCL team. His research goal was to produce the first middleware solution to integrate HPC and IoT. JCL code migrated to github, enabling tracking issues, improvements and bug loggins. A complete test planning to cover all API services was started and improved during 2016.


In 2017, JCL became a hybrid middleware solution, e.g., a solution for IoT applications, IoT with HPC requirements applications, HPC applications and for HPC with IoT requirements applications. Platforms with Java 8 and Android, but also the Arduino platform are now supported by JCL. The domain www.javacaela.org and the new JCL website are finally online. The team has now two high-school students, one master's degree student and eight undergraduate students. Several investigation themes started: i) JCL dashboard application to monitor JCL cluster visually; ii) JCL application for OLAPing over clusters composed of GPUs and CPUs things; iii) JCL Graph, a Java library designed for massive graph-storage and graph-processing; iv) JCL application for text classification; v) JCL application for financial analysis of historical news, vi) JCL optimization solver for huge instances of several combinatorial problems, and vii) JCL application for DNA, RNA and protein sequence alignments. Leonardo, the second master student in JCL, graduated in 2017. Estevão graduated in Computer Science undergraduate course. Two new papers about JCL were accepted in SBESC symposium and two other papers are under review in top-quality journals. A JCL proposal was submitted to Serrapilheira institution to raise funds for equipments and scholarships.


Two new investigations themes will start soon: i) a PhD student will investigate new scheduling algorithms for IoT and HPC technologies; and ii) a master student will investigate new fault tolerance algorithms for IoT and HPC technologies. In both scenarios JCL will be used as a test middleware platform. New API services for processing tasks will simplify the execution of distributed/parallel pipelines over a cluster or map/reduce tasks over a cluster. New IoT platforms, such as IoS and MicaZ, will be supported by JCL. Comparative evaluations against standards and market-leaders will continue to prove JCL efficiency, and social initiatives with kids will continue to prove JCL simplicity. Two new Journal papers (Qualis A2) has been published in 2018, one about JCL HPC and the other about JCL IoT/HPC integration. Two undergraduate students will graduate in 2018 using JCL as their investigation themes.