Commit Graph

5 Commits

Author SHA1 Message Date
Ivan Paljak
d3e00635c6 Measure read throughput in HA
Summary:
This macro benchmark measures read throughput in HA.
The test first creates a random graph with a given number of nodes
and edges. After that, it concurently performs the following query
for 10 seconds:

```
MATCH (n {id:$random_id})-[e]->(m) RETURN e, m;
```

In other words, it randomly picks a node and returns all its neighbours.

Locally measured results are as follows:

| nodes | edges | queries per second |
|  100  |  500  |        8900        |
| 1000  | 5000  |        2700        |
| 10000 | 50000 |        1200        |

Running the same test on Memgraph single node yields very similar results
(up to a few hundred queries).

Reviewers: msantl

Reviewed By: msantl

Subscribers: pullbot

Differential Revision: https://phabricator.memgraph.io/D1916
2019-03-12 14:55:04 +01:00
Ivan Paljak
276672c310 Fix flaky ha feature benchmark
Reviewers: msantl

Reviewed By: msantl

Subscribers: pullbot

Differential Revision: https://phabricator.memgraph.io/D1820
2019-01-21 14:37:38 +01:00
Matija Santl
49f906f384 Disable flaky HA benchmark
Summary:
It seems like the HA benchmark is flaky and unrelated builds fail
because of it. Lets disable it for now until we figure out what is happening.

Reviewers: ipaljak

Reviewed By: ipaljak

Subscribers: pullbot

Differential Revision: https://phabricator.memgraph.io/D1814
2019-01-17 12:54:30 +01:00
Matija Santl
c39a2278ae Add HA benchmark
Summary:
A simple benchmark that starts a HA cluster with 3 machines.
The benchmark issues only `CREATE (:Node)` queries.

Local results (debug build), for this raft config, are:
```
duration 4.26899
executed_writes 300
write_per_second 70.2743
```

Reviewers: ipaljak, mferencevic

Reviewed By: ipaljak

Subscribers: pullbot

Differential Revision: https://phabricator.memgraph.io/D1798
2019-01-15 13:32:04 +01:00
Matija Santl
57b84f2da3 Add kafka benchmark
Summary:
In order to add kafka benchmark, `memgraph_bolt.cpp` has been split.
Now we have `memgraph_init.cpp/hpp` files with common memgraph startup code.
Kafka benchmark implements a new `main` function that doesn't start a bolt
server, it just creates and starts a stream. Then it waits for the stream to
start consuming and measures the time it took to import the given number of
entries.

This benchmark is in a new folder, `feature_benchmark`, and so should any new
bechmark that measures performance of memgraphs features.

Reviewers: mferencevic, teon.banek, ipaljak, vkasljevic

Reviewed By: mferencevic, teon.banek

Subscribers: pullbot

Differential Revision: https://phabricator.memgraph.io/D1552
2018-08-29 16:35:31 +02:00