Senior Systems Administrator
Brighton, United Kingdom
Senior systems administrator for telecoms company. Responsible for deployment of SIP-based software for large telecoms customer. Duties include:• Automating operating system and software deployment across multiple customer sites(Cobbler, Puppet).• Packaging in-house developer code (RPM).• Deployment of scalable micro-service architecture (Node.js, Nginx, RabbitMQ).• High Availability services (Veritas Cluster Services, Heartbeat).• On-site monitoring and alarming (Nagios, Munin).• Writing and maintaining plugins and orchestration scripts written in Ruby, Python, Perland Bash to query many services/metrics using language libraries, curl (accessingRESTful APIs) and standard UNIX commands.• Responding to emergencies from on-call systems.• Code merge/release management (Subversion, Git). Maintaining Puppet code as a fully versioned product developed along SDLC stages.• Programmatically parsing, editing and generating config files (YAML, JSON, ini).• Issue tracking/documentation/revision control integration (JIRA/Confluence).• Storage (NFS, GlusterFS).• Deploy test environment to cloud (DigitalOcean). First writing ruby scripts to accessprovider APIs, then using VM creation tool (Vagrant).• Continually building product and running various lint parsers (Jenkins).• Staff training.Accomplishments:• Rationalizing Puppet code into Roles/Profiles pattern (recommended by PuppetLabs).• Reducing code base to one third by utilizing carefully selected Forge modules.• Migrating product code from Puppet 2.7 to 3.6, programmatically migrating customhierarchical Puppet node definitions of customer sites into Hiera YAML files. DesigningHiera structure to accommodate multi-OS installations and clearly disambiguateprogram data from customer data.• Migrating remaining single points of failure into HA framework (Heartbeat, GlusterFS,Veritas Cluster Service). Achieved long-term company target of having no singleserver failure impact service.