Skip to content

petrosze/ContHIFUN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ContHIFUN

Incremental Continuous Queries Evaluation in HIFUN
HIFUN is a high-level query language for expressing analytic queries over big data sets. In this project, based on HIFUN language we implement the corresponding algorithms for the incremental evaluation of continuous queries. We translate HIFUN queries to SQL and MapReduce tasks using Spark Streaming and Spark Structured Streaming, supporting also various queries rewriting methods proposed by HIFUN.

Datasets
A synthetic dataset used split into 10 files and pushed into a queue, treaded as a batch of data and processed like a stream. To distributed the data uniformly among all the cluster works, the data follows uniform distribution and stored in distributed file system. In case of the map reduce execution model the source data set is provided as a single text file and the analysis context used to extract values to formulate queries (in the form of triples as proposed of HIFUN theory), whereas in the case of SQL execution model, the source data set is provided as a structure dataset according to relational schema.

Queries
Common Grouping Measuring Rewriting Rule (CGMRR)
Queries used without rewritings
Q1 = (g1, m1, sum)
Q2 = (g1, m1, min)
Q3 = (g1, m1, max)
Q4 = (g1, m1, count)
Q5 = (g1, m1, avg)
Query after rewriting
Q = (g1, m1, {sum, min, max, count, avg})

Common Grouping Rewriting Rule (CGRR)
Queries used without rewriting
Q1 = (g1, m1, sum)
Q2 = (g1, m2, min)
Q3 = (g1, m3, max)
Q4 = (g1, m4, count)
Q5 = (g1, m5, avg)

Query after rewriting
Q = (g1, {m1, sum}, {m2, min}, {m3, max}, {m5, avg})

Common Measuring and Operation Rewriting Rule (CMORR)
Queries used without rewriting
Q1 = (g1, m1, sum)
Q2 = (g2, m1, sum)
Q3 = (g3, m1, sum)
Q4 = (g4, m1, sum)
Q5 = (g5, m1, sum)

Queries after rewritings
Q1 = ((g1 ^ g2 ^ g3 ^ g4 ^ g5), m1, sum)
Q2 = (projG1, Q1, sum)
Q3 = (projG2, Q1, sum)
Q4 = (projG3, Q1, sum)
Q5 = (projG4, Q1, sum)
Q6 = (projG5, Q1, sum)

Basic Rewriting Rule (BRR)
Queries used without rewriting
Q1 = (g11°g1, m1, sum)
Q2 = (g12°g1, m1, sum)
Q3 = (g13°g1, m1, sum)
Q4 = (g14°g1, m1, sum)
Q5 = (g15°g1, m1, sum)

Queries after rewriting
Q1 = (g1, m, sum)
Q2 = (g11, Q1, sum)
Q3 = (g12, Q1, sum)
Q4 = (g13, Q1, sum)
Q5 = (g14, Q1, sum)
Q6 = (g15, Q1, sum)

Launching the project with spark-submit
Once an experiment configuration is bundled, it can be launched using the bin/spark-summit script and runs on Mesos cluster. This script takes care of setting up the classpath with Spark and its dependencies and can support different cluster configutations and deploy modes.

spark-submit \
--class src.main.java.evaluator.mapreduce.MainRewr \
--driver-memory 20G \
--executor-memory 200G \
--conf spark.speculation=true \
--master mesos://zk://clusternode5:2181,clusternode6:2181,clusternode7:2181/mesos \
hifun.jar

References
[1] Spyratos, N.; Sugibuchi, T. A High Level Query Language for Big Data Analytics. 2014.
[2] Spyratos, N.; Sugibuchi, T. HIFUN - a high level functional query language for big data analytics. Journal of Intelligent Information Systems 2018,51, 529–555.718
[3] Zervoudakis, P.; Kondylakis, H.; Plexousakis, D.; Spyratos, N. Incremental Evaluation of Continuous Analytic Queries in HIFUN. ISIP, 2019, pp. 53–67.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages