tursom/memgraph
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
b92b8c555a
memgraph/tests/public_benchmark/ldbc
History
florijan b92b8c555a Fix LDBC durability 
Reviewers: teon.banek

Reviewed By: teon.banek

Subscribers: pullbot

Differential Revision: https://phabricator.memgraph.io/D1001
2017-11-23 11:02:54 +01:00
..
config Cleanup durability config, docs, CHANGELOG 2017-11-21 10:17:13 +01:00
plots A python script that visualizes latency results from LDBC. 2017-09-04 16:36:42 +02:00
results A python script that visualizes latency results from LDBC. 2017-09-04 16:36:42 +02:00
test_cases Add filtering BFS by edge type 2017-09-12 11:29:38 +02:00
.gitignore LDBC: Integrate with Apollo 2017-09-21 22:29:42 +02:00
build_dataset Fix LDBC durability 2017-11-23 11:02:54 +01:00
cleanup LDBC: Integrate with Apollo 2017-09-21 22:29:42 +02:00
continuous_integration Added operation count to LDBC update runs. 2017-09-24 23:02:12 +02:00
index_creation.py Integrated C++ CSV to snapshot into LDBC. 2017-09-23 23:12:23 +02:00
ldbc-snb-impls-reads.properties Integrated C++ CSV to snapshot into LDBC. 2017-09-23 23:12:23 +02:00
ldbc-snb-impls-updates.properties ldbc: Separate running short reads from updates 2017-09-07 10:45:09 +02:00
README.md Integrated C++ CSV to snapshot into LDBC. 2017-09-23 23:12:23 +02:00
requirements.txt LDBC: Integrate with Apollo 2017-09-21 22:29:42 +02:00
run_benchmark Cleanup durability config, docs, CHANGELOG 2017-11-21 10:17:13 +01:00
setup Integrated C++ CSV to snapshot into LDBC. 2017-09-23 23:12:23 +02:00

README.md

LDBC benchmarks

How to run the benchmark against Neo4j OR Memgraph?

cd memgraph/tests/public_benchmark/ldbc
./setup
./build_dataset [--scale-factor 1]
# To run short reads by default, just call:
./run_benchmark --create-index --run-db memgraph # or neo4j
# To run update queries pass the properties file for updates and slow down
# the execution by setting a larger time compression ratio.
./run_benchmark --create-index --run-db memgraph --test-type updates \
                --time-compression-ratio 1.5

How to run a specific test?

cd memgraph/tests/public_benchmark/ldbc/ldbc-snb-impls/snb-interactive-neo4j
mvn exec:java -Dexec.mainClass="net.ellitron.ldbcsnbimpls.interactive.neo4j.util.QueryTester" -Dexec.args="shortquery1 933 --repeat=10000 --timeUnits=MICROSECONDS"

Useful terminal commands

How to find out Default Query Limits?

cd memgraph/tests/public_benchmark/ldbc/ldbc_driver
ag "DEFAULT_LIMIT"

How to run test cases?

cd memgraph/tests/public_benchmark/ldbc
neo4j-client --insecure -u "" -p "" -o test_cases/results/short_query_2.oc.out -i test_cases/queries/short_query_2.oc localhost 7687

How to create indexes manually?

cd memgraph/tests/public_benchmark/ldbc
source ve3/bin/activate
./index_creation.py ldbc-snb-impls/snb-interactive-neo4j/scripts/indexCreation.neo4j

Where is and how to use LDBC plotting?

cd memgraph/tests/public_benchmark/ldbc
source ve2/bin/activate
cd memgraph/tests/public_benchmark/ldbc/ldbc_driver/plotting
python make_charts_all_queries_all_metrics.py /path/to/ldbc_driver/results/LDBC-results.json legend_location

TODOs

  1. Run & validate all Short queries + optimize MemGraph performance.
  2. Write the log parser.
  3. Visualize the results.
  4. Run & validate all Update queries + optimize MemGraph performance.
  5. Run apacaci code an compare the results (Do We Need Specialized Graph Database?).
  6. Optimize MemGraph performance.
  7. Scale benchmarks (scale factors 1, 3, 10).
  8. Test against Postgres and make comparison between Neo, Memgraph & Postgres. Latency + Throughput.
  9. Run & validate all queries + optimize MemGraph performance.
  10. Optimize MemGraph spinup time.