Skip to content
pewsou edited this page Nov 4, 2017 · 11 revisions

Welcome to SCUD - the toolkit implementing scheduling functionality under Generalized Processor Sharing paradigm. This software provides building blocks for routing chain of custom defined objects. The blocks include but not limited to: generalized Queue, Dropper (filter) and Scheduler. Combining these entities one can insert his data objects into FIFO containers and then select some container each time and pull data object. Objects may be dropped (filtered out).

Version 0.1.13 is available.

Q: When can we need it?

A: Consider next scenario: you have in your SW some concurrent threads which post SQL queries to some built-in database facility. Only one query can be processed by database at the time. So at any time point there is need to select from available queries strictly one and pass it to database. For this purpose it is suitable to provide some queuing ability (to collect yet unprocessed queries), some scheduling facility - to select at each moment the query to process, and some filtering facility - to sort out excessive queries. All these are provided by this toolkit. Another example: data packets coming from communication facility may be routed and filtered in the same manner.

Q: What Scud does with my data objects?

A: Nothing but passing them along the rout. The developer is responsible for allocation, destruction and other forms of management of these objects.

Coming soon: more documentation

Clone this wiki locally