2016-11-25 12:34:08 +02:00
|
|
|
# leak
|
|
|
|
|
|
|
|
|
|
- 1. Spawn a bunch of servers:
|
|
|
|
|
- another-fast: a node with a linear memory leak
|
|
|
|
|
- fast: a node with a linear memory leak
|
|
|
|
|
- slow: a node with a memory leak that grows very slowly
|
|
|
|
|
- plain: a node with no memory leak
|
|
|
|
|
- 2. Spawn an [artillery](https://artillery.io) for each node that loads it with a small but constant stream of requests
|
|
|
|
|
- 3. Spawn Prometheus that watches the cpu/memory of each node
|
|
|
|
|
|
|
|
|
|
Then, locally we start the same server and we can see the different instances and an aggregate of the metrics for each job.
|
|
|
|
|
|
|
|
|
|
## usage
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
λ docker-compose up
|
|
|
|
|
λ node .
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
Go to http://127.0.0.1:8000/ and see the result.
|
|
|
|
|
The [Prometheus](https://prometheus.io) is also listening at http://127.0.0.1:9090/
|
|
|
|
|
|
|
|
|
|
## example
|
|
|
|
|
|
|
|
|
|
|
