Skip to main content

Zeebe ChaosChaos Summaries

All posts

2025

Building Confidence at Scale: How Camunda Ensures Platform Reliability Through Continuous Testing
Stress testing Camunda
Testing retention of historical PIs in Camunda 8.8
Resilience of dynamic scaling
REST API: From ForkJoin to a Dedicated Thread Pool
Resiliency against ELS unavailability
Dynamic Scaling: probing linear scalability
Follow up REST API performance
Performance of REST API
How does Zeebe behave with NFS
Lower memory consumption of Camunda deployment

2024

News from Camunda Exporter project
Impact of Camunda Exporter on processing performance
Camunda Exporter MVP
Optimizing cluster sizing using a real world benchmark
Improve Operate import latency
Operate load handling
Using flow control to handle bottleneck on exporting
Using flow control to handle uncontrolled process loops
Reducing the job activation delay

2023

Broker Scaling and Performance
Dynamic Scaling with Dataloss
Dynamically scaling brokers
Job push resiliency
Job push overloading
Hot backups impact on processing
Using Large Multi-Instance
Continuing SST Partitioning toggle
SST Partitioning toggle
Gateway Termination
Recursive call activity

2022

Message Correlation after Network Partition
Bring Deployment distribution experiment back
Standalone Gateway in CCSaaS
High Snapshot Frequency
Handling of Big Variables

2021

Worker count should not impact performance
Not produce duplicate Keys
Throughput on big state
Recovery (Fail Over) time
Old-Clients
Slow Network
Full Disk Recovery
Time travel Experiment
Corrupted Snapshot Experiment Investigation
BPMN meets Chaos Engineering
Set file immutable
Camunda Cloud network partition
Fault-tolerant processing of process instances
Automating Deployment Distribution Chaos Experiment
Deployment Distribution
Network partitions
Disconnect Leader and one Follower

2020

Message Correlation after Failover
Many Job Timeouts
Investigate failing Chaos Tests
Non-graceful Shutdown Broker
Gateway memory consumption
Multiple Leader Changes
Play around with ToxiProxy
Experiment with Camunda Cloud
Experiment with Low Load
Experiment without Exporters
Big Multi Instance
Experiment with Timers and Huge Variables
Extract K8 resources from namespace
Gateway Network Partition
Correlate Message after failover
High CPU load on Standalone Gateway
First Chaos Day!

Tags

A

availability38

C

chaos1

D

data6

P

performance12

R

reliability1
resiliency2

S

scalability1
stress-testing1

T

tests2
tools2

Community

Stack Overflow
Forum
Twitter

More

GitHub

Copyright © 2025 Zeebe Chaos. Built with Docusaurus.