Session Name: CI in Linux Kernel filesystem development with continuous performance analysis
"My plan is to cover the presentation based on the below four key points:
1. General discussion about continuous integration (CI) in kernel development:
- This discusses why CI is needed and how CIs are implemented in linux kernel development
2. Why performance is important and how often we test performance
- This section discusses the importantance and impacts of performance of kernel modules
3. How we process performance results with elastic (Open source version) and generate alerts (also open) based on performance improvement or degrade
- A simple explaination of how we use Elasticsearch for storing performance results and plotting performance results with Kibana.
Also discuss how alerts (Elastalerts) are implemented and notified via emails and in chat.
4. Future plans and improvements
- How Elastic's Machine learning (anomaly detection) can process performance results"
I am a DevOps lead at Tuxera and an avid open source contributor with over 15 years experience in various positions in big and small sized companies. At Tuxera Oy, I implemented CI from ground up with open source tools and home brewed test automation framework. In the past, I have implementing container storage interface (MooseFS CSI) for Kubernetes on distributed on-cloud/on-premise servers. Recently, my focus has been trying to make Blockchain technologies especially Linux Foundation's Hyperledger Fabric usable for companies. I hold a Master’s in Mobile Technology from University of Jyväskylä, Finland."