Hi Ratnadeep, Thanks for your interest in the topic. I would look first at JobCapacityController, and give the Hyracks paper(https://asterix.ics.uci.edu/pub/Hyracks.pdf) a read to get a general sense of the broad memory constraints and controls for each Hyracks job. Then, you can look at SetMemoryRequirementsRule to see how Asterix is currently calculating the required memory for a job during compile time (the compiled form of a query)
A good first contribution towards this topic might be some more tests on the Asterix side regarding the current functionality. There are some tests in Hyracks for the CapacityController, but to my knowledge there aren't many that check the behavior of the parameters that are used on the Asterix side. Thanks, - Ian On Mon, Mar 2, 2026 at 11:14 AM Ratnadeep Mohanta <[email protected]> wrote: > > Hi everyone, > > I’m Ratnadeep from India, currently a B.Tech CSE student at Amity > University. I have been programming for around two years and have worked > with languages such as C, C++, and Java. I am also currently working in web > development and have experience building full-stack websites. > > I’ve recently started exploring the dynamic memory management idea in > Apache AsterixDB, particularly around improving the current static > budgeting approach for memory-intensive operators. I’ve begun going through > the documentation and relevant parts of the codebase to understand how > operator memory is currently allocated and managed. > > I’m very interested in contributing further in this area and would really > appreciate guidance on what I can work on next to make meaningful > contributions. Are there specific components, issues, or discussions I > should focus on? > > Thanks, > Ratnadeep
