Skip to content

Latest commit

 

History

History
47 lines (40 loc) · 3.54 KB

README.markdown

File metadata and controls

47 lines (40 loc) · 3.54 KB

![alt text][logo] [logo]: https://raw.githubusercontent.com/epfldata/squall/master/resources/graphics/logo.jpg "Logo Title Text 2"

#Squall Squall is an online query processing engine built on top of Storm. Similar to how Hive provides SQL syntax on top of Hadoop for doing batch processing, Squall executes SQL queries on top of Storm for doing online processing. Squall supports a wide class of SQL analytics ranging from simple aggregations to more advanced UDF join predicates and adaptive rebalancing of load. It is being actively developed by several contributors from the EPFL DATA lab. Squall is undergoing a continuous process of development, currently it supports the following:

  • SQL (Select-Project-Join) query processing over continuous streams of data.
  • Full fledged & full-history stateful computation essential for approximate query processing, e.g. Online Aggregation.
  • Time based Window Semantics for infinite data streams (currently in-progress).
  • Theta Joins: complex join predicates, including inequality, band, and arbitrary UDF join predicates. This gives a more comprehensive support and flexibility to data analytics. For example, Hive plans to support theta joins in response to user requests.
  • Continuous load balance and adaptation to data skew.
  • Usage: An API for arbitrary SQL query processing or a frontend query processor that parses SQL to a storm topology.
  • Throughput rates of upto Millions of tuples/second and latencies of milliseconds on a 16 machine cluster. Scalable to large cluster settings.
  • Out-of-Core Processing: Can operate efficiently under limited memory resources through efficient disk based datastructures and indexes.
  • Guarantees: At least-once or at most-once semantics. No support for exactly-once semantics yet, however it is planned for.
  • Elasticity: Scaling out according to the load.
  • DashBoard: Integrating support for real time visualizations.

Example:

Consider the following SQL query:

SELECT C_MKTSEGMENT, COUNT(O_ORDERKEY)
FROM CUSTOMER join ORDERS on C_CUSTKEY = O_CUSTKEY
GROUP BY C_MKTSEGMENT

Through the Squall API, the online distributed query plan (full code) can be simply formulated as follows:

Component relationCustomer = _queryBuilder.createDataSource("customer", conf)
                                          .add(new ProjectOperator(0, 6))
                                          .setOutputPartKey(0);
Component relationOrders  = _queryBuilder.createDataSource("orders", conf)
                                          .add(new ProjectOperator(1))
                                          .setOutputPartKey(0);
_queryBuilder.createEquiJoin(relationCustomer, relationOrders)
                                          .add(new AggregateCountOperator(conf)
                                          .setGroupByColumns(1));

Documentation

Detailed documentation can be found on the Squall wiki.

Contributing to Squall

We'd love to have your help in making Squall better. If you're interested, please communicate with us your suggestions and get your name to the Contributors list.

License

Squall is licensed under Apache License v2.0.