Experienced Systems Administrator
| Address: | I live in Worthing, West Sussex | |
| Email: | website@incomprehensible.co.uk | |
Tim Clark
DevOps Engineer
... and Dad
The very simple homepage of Tim Clark
Tim is a DevOPS-enabled Operations Engineer with over 20 years of well-rounded experience and a degree in the Computer Science
Tim particularly enjoys building and creating highly observable container based services
Tim has a fantastic family and loves gaming, (watching) football, podcasts and many things computer based
Tim has helped lead Breakfast Church (in various forms) at Maybridge with a great team of people for most of 20 years
Tim has developed a worrying tendency to write in the third person while producing this website
Contact Details
General Skills
Containerisation Projects
Linux based technology
Encouraging DevOps adoption
Cloud Services and migrations
Building OPS teams
Web Design
Skills
Container technology
Docker Docker Swarm Traefik KubernetesObservability
Grafana Nagios NewRelic Prometheus JaegerOps skills
Apache Haproxy Linux Mysql Varnish Vmware Drupal Java Kafka MS SQL Postgres Solr ZookeeperCI/CD
Jenkins Jenkins Pipeline Git GitHub GitLabCloud Services
Google CloudBuild Google CloudRun AWS Elasticbeanstalk AWS CloudFormation AWS EC2 AWS RDSConfig Management / Provisioning
Ansible Chef Puppet Terraform AWS Cloudformation ElasticbeanstalkThe journey to so far...
Pinnacle Internet Homepage ISP Hosting
Pinnacle were one of the first Business to Business ISPs in the country. I installed their first datacentre based servers - and quickly became the Senior Systems Administrator. Pinnacle were bought out by Precedent Communications in 2003.
Precedent Homepage ISP Hosting
“Meaningful ideas, intelligently delivered”
Precedent specialised in strategy, web, brand and mobile.
I grew the Hosting offering into profitable multi-datacentre, geographically disparate service
Our customers included British Medical Association, Parkinsons UK, Wildlife Trust, British Heart Foundation and Anthony Nolan
I managed a global team who supported Precedent systems - while helping to architect HA and load balanced solutions for customers.
I oversaw the hosting platform and architecture, incident management and budget along with managing the OPS team spread around the UK and Australia.
Semantico / Highwire Homepage Publishing DevOps
I joined Semantico Ltd just as they were purchased by Highwire Press - working on publishing platforms and helping lead on all things DevOps.
I introduced DevOps concepts and tooling to a largely traditional environment which contained a scattering of 'unicorn' projects, leading to fragmentation.
MPS Limited Homepage Publishing Hosting
The next step of the journey started in 2020 - when India based MPS purchased Highwire
New challenges include intregrating with a largely India based company - and consolidating all existing hosting into the cloud
Achievements
"I don't like blowing my own trumpet." - George
"You might at least told us you had a trumpet." - Blackadder
Converted Legacy Systems to DockerMPS Highwire
I moved around 100 Java and Go legacy applications onto Docker with a mixture of Jenkins, Swarm and Ansible - building in centralised logging, OpenTracing and other DevOps favourites while we were doing it.Introduced DevOps strategy MPS Highwire
Empowered Developers to deploy their own code by building CI/CD frameworks. Worked alongside teams to convert their processes to align with modern standards. Introduced Pipelines, Centralised logging, Containers, Observability (Prometheus, NewRelic, Opentracing).Building a team of Systems/Devops staff across geographical boundaries MPS Highwire Precedent
I build OPS teams spanning Cardiff/Edinburgh/Perth/Melbourne/Worthing/London at Precedent and DevOps teams spanning Belfast, Brighton, Los Gatos USA and India for Highwire.Config Management overhaul MPS HighwirePrecedent
At Precedent I introduced Puppet to manage the whole server estate. At Highwire I was part of a three man team that introduced Ansible as a replacement from the ageing CFE2,CFE3 and Puppet 2 infrastructure. We provisioned VMware machines on-premises with this system - and I taught some excellent junior staff how to manage it day to day.MigrationsMPS HighwirePrecedent
Along with the team at Highwire, we executed a move of 1000's of VMs from a Datacentre at Stanford University to a newly specced Virtual environment with a US based Cloud provider. Immediately after that was completed we started a similar consolidation project from our UK based DC to the US. Lots of late nights - but the end result was a tiny number of customers even noticing the change!Overengineering
I'm always recommending the simplest possible solution. This site is the exception that proves the rule - it is horribly overengineered
Github latest
Github has the code committed - there are two branches at the moment - 'main' and 'dev'
By storing the code in Git we have accountability, reproducability and a 'last known good' position
CloudBuild Docs
Google Cloudbuild is watching the 'main' branch.
As soon as it spots a change it will automatically build a new container and store it in GCR. Using the Dockerfile instruction in the repo.
By building it in Docker - we ensure reproducability across environments
CloudRun Docs
CloudRun then picks up the built container and deploys it out to https://www.incomprehensible.co.uk/
This is all running on the free tier!
Analytics Docs
Javascript code in the header of the HTML will report usage stats to Google Analytics
All still free!
Logs (Google Operations) Docs
Logs are pushed into Google Operations - for easy analysis
Also free!
Alerting Docs
Alerting will watch the homepage for the string "Tim Clark" and whether the SSL is working and will alert if it isn't