Brandenburg, Germany · DESY Zeuthen — Scientific Computing
Ingo Ebel
Open Source Infrastructure · Scientific Computing · Tech Journalism
Linux · Ceph · Proxmox · Nextcloud · GitLab · Networking · Automation
Building and operating reliable open-source infrastructure — and talking about the technology behind it.
![Portrait of Ingo Ebel [PLACEHOLDER: photo not yet uploaded]](/assets/images/ingo-ebel.jpg)
01 About
I hold a Master of Science in Medieninformatik and have spent more than two decades with Linux, open source and infrastructure I run myself. What started as curiosity about how systems actually work became the way I work: close to the machine, close to the protocol, close to the logs.
Today I work in Scientific Computing at DESY Zeuthen, where I administer Linux, storage and compute infrastructure and look after GitLab as a platform for research and engineering teams. The work ranges from day-to-day operations to migrations, capacity planning and the kind of troubleshooting that starts with a symptom and ends three layers further down.
Alongside that I run open-source based IT services and infrastructure for companies and organisations through JIT Creatives — collaboration platforms, mail, storage, networking and the automation that keeps them reproducible. Ansible and CI pipelines are not a hobby here; they are how infrastructure stays maintainable over years.
The second half of my work is communication. For many years I have been active as a tech journalist, podcaster and interviewer, reporting on Linux, open source, cloud native, digital sovereignty and self-hosting, and talking with the people who build these technologies. Passing on knowledge — in interviews, in IT training, in documentation — is part of the same job.
02 Infrastructure & Expertise
Not a list of tools I have read about — these are the areas I build, operate, migrate and debug.
-
Linux & Systems
Operating and troubleshooting complex Linux infrastructures: automation, package management, services and system integration.
-
Storage & Virtualization
Building and running distributed storage and virtualisation platforms with a focus on high availability, performance and fault analysis.
-
Collaboration Infrastructure
Running highly available open-source collaboration platforms together with their database, cache, authentication and proxy infrastructure.
-
GitLab & Automation
GitLab administration and the automation of deployment and infrastructure processes.
-
Mail Infrastructure
Operating self-hosted mail infrastructure including spam filtering, mailing lists and DNS integration.
-
Networking
Planning, operating and troubleshooting segmented networks and enterprise switching infrastructure.
-
Monitoring
Monitoring and analysis of server, network and infrastructure metrics.
-
Cloud Native
Operating, architecting and assessing cloud-native technology — with an eye on what actually holds up in production.
03 Selected Infrastructure Work
Descriptions are deliberately kept at architecture level. No internal hostnames, addresses, credentials or customer data.
-
High-Availability Nextcloud Infrastructure
Design and operation of a highly available Nextcloud platform for several hundred users: shared storage on CephFS, a replicated PostgreSQL backend managed with Patroni, a Redis cluster for locking and sessions, Traefik in front and single sign-on through Keycloak/OIDC. Upgrades, configuration and rollout run through GitLab CI/CD, with metrics and alerting on top.
- High availability
- Shared storage
- Authentication
- Upgrade strategy
- Automation
- Monitoring
-
Distributed Ceph Storage
Operating and troubleshooting distributed Ceph storage for virtual machines, file services and compute environments — from pool and placement layout over 10 GbE fabrics to the slow-ops and recovery analysis that keeps a cluster healthy under real load.
- RBD & CephFS
- Capacity
- Recovery
- Fault analysis
-
GitLab Infrastructure & Automation
Administration and continued development of self-hosted GitLab systems, together with the automation of infrastructure and deployment processes: runners, CI pipelines, Ansible roles and AWX as the execution layer.
- Platform operation
- CI/CD
- Infrastructure as code
-
Open Source Mail Infrastructure
Operating a self-hosted mail platform: Postfix as MTA, Dovecot for delivery and access, Rspamd with Redis for filtering, Mailman for mailing lists, and the DNS records that make all of it verifiable from the outside.
- Deliverability
- Spam filtering
- Mailing lists
- DNS
04 Open Source & Community
Open source is not a technology choice I made at some point — it is the environment I have worked in for more than twenty years. The same software I run in production is the software I report on, and the communities behind it are the ones I talk to.
That combination shapes how I look at infrastructure: what can you actually operate yourself, what does it cost in attention, where does a project stay dependable over years, and what does digital sovereignty mean once you are the one carrying the pager. Conferences and community events are where a lot of that gets discussed in the open — and where most of my interviews happen.
- Linux
- Open Source
- Self-Hosting
- Digital Sovereignty
- Open Infrastructure
- Cloud Native
- Community Events
- Knowledge Transfer
05 Podcasts & Tech Journalism
Talking with the people who build open-source technology.
-
RadioTux
German-language media project covering Linux, open source and net culture. I work there as host, interviewer and technology journalist.
-
Binärgewitter
Podcast on open source, software development, infrastructure, security, AI and net culture.
-
Nextcloud Podcast
Interviews and conversations with developers, community members and decision-makers from the Nextcloud ecosystem — development, app ecosystem, open source, digital sovereignty, collaboration and infrastructure.
Episodes
[Podcast episode data]
Episodes are maintained in js/data.js and appear here automatically.
Conference Interviews
Interviews and reports from open-source events such as KubeCon, FrOSCon, Chemnitzer Linux-Tage, Nextcloud Conference and Cephalocon.
- KubeCon
- FrOSCon
- Chemnitzer Linux-Tage
- Nextcloud Conference
- Cephalocon
[Podcast episode data]
Conference interviews are maintained in js/data.js.
As a guest
The other way round — conversations in which I was the one being asked.
[Podcast episode data]
Guest appearances are maintained in js/data.js.
06 Articles, Interviews & Talks
Written pieces, interviews and talks on infrastructure and open source.
[Article data]
Articles are maintained in js/data.js and appear here automatically.
Talks
[Talk data]
Talks are maintained in js/data.js.
Press & mentions
[Press data]
Mentions are maintained in js/data.js.
07 Lab & Smart Infrastructure
Exploring how open protocols, local automation and energy systems can work together without depending entirely on proprietary cloud services.
The same questions as in the data centre, one scale down: where does state live, what happens when the uplink is gone, how do you get metrics out of a device that was never meant to give you any, and what does it take to keep a building automation stack maintainable for a decade.
- Home Assistant
- KNX
- Modbus
- MQTT
- Sensors
- Photovoltaics
- Energy Management
- Building Automation
08 Contact
Interested in open source, infrastructure or a technical conversation?