S-Store Transactional Streaming Data Management System
Java C++ C Python Shell Makefile Other
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
.metadata
properties
scripts
src
tests
third_party
tools
.gitignore
COPYING
Dockerfile
LICENSES
README.md
build-common.xml
build.py
build.xml
buildtools.py
hstore
inputclient
log4j.properties
out-11.ogv
replace.py
runinputclient
runsstorev1.sh
sparkbatchvoter
sstore
stream-generator.jar
version.txt
votefilegenerator
votegenclient
votegenserver

README.md

S-Store

S-Store is an experimental system for streaming data management with shared mutable state. It is a hybrid streaming system with a main-memory OLTP data management system base. It integrates transaction processing with push-based semantics.

More information and documentation is available at: http://sstore.cs.brown.edu

Supported Platforms

S-Store is built on [H-Store]http://hstore.cs.brown.edu, and is known to work on the following platforms. Please note that it will not compile on 32-bit systems.

  • Ubuntu Linux 9.10+ (64-bit)
  • Red Hat Enterprise Linux 5.5 (64-bit)

Dependencies

Quick Start

  1. First build the entire distribution:

     ant build
    
  2. Next make the project jar file for the target benchmark. S-Store includes several benchmarks that are built-in and ready to execute. A project jar contains all the of stored procedures and statements for the target benchmark, as well as the cluster configuration for the database system.

     export SSTORE_BENCHMARK=votersstoreexample
     ant sstore-prepare -Dproject=$SSTORE_BENCHMARK
    
  3. You can now execute the benchmark locally on your machine with two partitions

     ant sstore-benchmark -Dproject=$SSTORE_BENCHMARK
    

More information is available here: http://sstore-doc.readthedocs.io/en/latest/deploy.html