https://github.com/apetenchea/replicated-states-chaos

Replicated States Stress Testing for Arangodb
https://github.com/apetenchea/replicated-states-chaos

Last synced: about 22 hours ago
JSON representation

Replicated States Stress Testing for Arangodb

• Host: GitHub
• URL: https://github.com/apetenchea/replicated-states-chaos
• Owner: apetenchea
• Created: 2022-05-17T18:36:10.000Z (over 2 years ago)
• Default Branch: master
• Last Pushed: 2022-07-28T14:10:03.000Z (over 2 years ago)
• Last Synced: 2023-08-27T19:28:39.356Z (about 1 year ago)
• Language: Python
• Size: 21.5 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: readme.md

Awesome Lists containing this project

README

        
Replicated States Stress Testing for ArangoDB

============================================

**Replicated states HTTP API might have changed**

In order to run stress testing on the prototype replicated state,

start a cluster with at least 12 DB servers and 3 coordinators (recommended).  

Run `chaos.py` and follow the logs in the console.

The files `actual.json` and `expected.json` represent the actual snapshot gathered form the coordinator and the expected final

state which was reconstructed.  

There can be a lot of output produced, hence you might want to redirect everything to a file: `python chaos.py &> log.txt`  

How I generally do it:

```

./scripts/startLocalCluster -a 1 -c 3 -d 12

python chaos.py -op 5000 &> log.txt

```

Then follow the logs in another terminal:

```

tail -f log.txt

```

When you want to re-run the whole thing, don't forget to shutdown the previous cluster.  

For debugging purposes, the contents of the replicated log are dumped into `replicated-log.json`.