Panagiotis Tsafaridis partly available

Panagiotis Tsafaridis


partly available
Profileimage by Panagiotis Tsafaridis Freelancer from Norderstedt
  • 22848 Norderstedt Freelancer in
  • Graduation: not provided
  • Hourly-/Daily rates:
  • Languages: German (Native or Bilingual) | English (Full Professional) | Greek (Full Professional)
  • Last update: 22.03.2021
Profileimage by Panagiotis Tsafaridis Freelancer from Norderstedt

You need an account to view this information.

Linux, Ansible, Saltstack, Puppet, Consul, Consul-Template, Terraform, Git, Gitlab, Jenkins, Proxmox, KVM, LXC, Docker, Kubernetes, Filesystems, Hadoop, Kafka, Cloudera, Bash, Graylog, Elasticsearch, Prometheus, Grafana, DataDog, Cassandra, MySQL, Apache, Nginx, Amazon AWS, Microsoft Azure, Cloudflare, Kitchen-CI, Inspec, Golang, Agile, Scrum, Kanban
  • 08/2020 - Present

    • xChange Solutions GmbH
    • 50-250 employees
    • Other
  • Freelance DevOps Engineer
  • Analysis and evaluation of the currently used infrastructure in AWS and software components to reduce costs and possible security risks.

    Successfully analyzed weaknesses of the Elsticsearch cluster and provided immediate change proposals to increase stability and ensure data integrity.

    Presentation of a more efficient AWS cloud architecture where self hosted services are replaced by cloud services and environments are isolated by different AWS organizations.

    Creation of guidelines for the use and development of Terraform automation scripts.

  • 03/2020 - Present

    • kodilabs GmbH
    • < 10 employees
    • Internet and Information Technology
  • Freelance Information Architect
  • Responsible for the development of a fully automated data center infrastructure architecture with "Ansible" for different web systems.

    Implementation of a central monitoring and logging system based on DataDog. Adding custom checks for own developments and setting up notification at different levels of criticality.

    Planning and automated implementation of patch management to quickly install security updates on all Linux servers.

    Implementation of various security concepts to protect web servers against attacks.

  • 08/2020 - 08/2020

    • Hege Consult UG
    • < 10 employees
    • Internet and Information Technology
  • Freelance DevOps Consultant
  • Project responsibility for the planning, conception and implementation of a hybrid-cloud network solution using Microsoft Azure and Terraform. The following Azure resources were used: Express Route, Route Tables, Firewall Manager, Application Gateway, VPN and the HUB-SPOKE network topology


    Designing a secure data traffic to manage the traffic between the data center, cloud and the public endpoints in a controlled way.

  • 02/2020 - 06/2020

    • Parship GROUP
    • 50-250 employees
    • Other
  • IT Consultant DevOps
  • Delivered a Proof-of-Concept for Puppet, which enabled the Operations Team to test changes on their local machines before they are pushed to production through Jenkins Pipelines, increasing development velocity and failure tolerance.   Improved the existing Infrastructure as Code solution for Cloudflare, based on Terraform, by redesigning the code architecture to a modular basis, which minimized duplication and allowed reuse of Cloudflare resources. Created a golang program leveraging the Cloudflare API to dump manually created entries, converting them into a format importable into Terraform state files. These changes improved maintainability and usage of the Terraform runtime.   Implemented DNS Zone Forwarding to multiple Environments for services registered in environment-specific Consul clusters, making it possible for developers to use the Consul DNS to browse their environment-specific Services by knowing only the datacenter and service names.   Updated expired Intermediate CA Certificates on all Puppet-managed hosts and added the new certificates to the system and browser truststores in an automated way.   Planned capacity for a future Kubernetes Topology in accordance with the existing services and resources used on hypervisors.

  • 11/2016 - 02/2020

    • Deposit Solutions GmbH
  • Senior Infrastructure Engineer
  • Played a key role in the main automation project converting the infrastructure from a mainly manual
    maintenance state into a completed automated state using Saltstack and GitLab, joining the design
    sessions and being responsible for a sizeable part of the implementation effort, later taking over the
    responsibility for being the main driver of automation and testing for any expansions using tools like
    kitchen and InSpec and respecting a git-flow model, ensuring that a machine park with more than
    300 hosts is completely managed through an automated and audited interface.

    Successful proof-of-concept of a monitoring solution for MySQL databases based on Prometheus,
    Grafana, and related exporters. This solution enabled developers to find inconsistencies amongst
    database schemas and Infrastructure engineers to adjust and enhance the database performance.

    Creation of a new segregated environment standard to run the company's applications, covering
    different roles such as web servers, backend servers, and batch job servers backed by a set of
    reusable Ansible deployment scripts.

    Successful migration of the company hypervisor technology from VMWare to Proxmox, greatly
    enhancing the automation capabilities of the platform and making it possible to run applications
    with LXC container technology.

    Design and implementation of a new load balancing architecture for dynamic request routing based
    on consul, consul-template, and nginx. This new architecture enabled central log aggregation,
    TLS-secured connections and dynamic routes based on service registration information stored from
    the Consul service catalog.

    Development of a new set of guidelines for service registration and discovery in the company,
    together with automation systems for Consul and Consul Template management using SaltStack,
    including enhancements to the core states and modules for Consul.

    Rollout of a new Docker Registry based on Gitlab EE together with an S3 backend with controlled
    access and encryption using in-house certificates. Provided documentation and examples for the
    internal teams to migrate to the new solution.

    Setup of a new test infrastructure based on Gitlab CI and dynamic build workers on AWS governed by
    auto scaling runners. The solution enabled all teams working in both infrastructure and application
    development to have an elastic environment for functional and end-to-end testing whilst greatly
    reducing the need for on-premise testing infrastructure, resulting in significant cost savings. The
    rollout of the solution included pair sessions with the development teams in order to adapt the
    solution to be compatible with semantic versioning and the Java build process used in the company.

    Successful PoC of infrastructure management on AWS using Terraform and a set of custom
    scripts, allowing better view on what would be required to migrate all AWS resources to a
    Terraform-managed setup.

    Creation of a team vision and moderation of discussion sessions which enabled the team to map
    pending work in form of a comprehensive roadmap, which is used to guide the Scrum planing

    Nginx Rollout of Graylog as the default log aggregator within the company, starting with PoC projects all
    Amazon AWS the way to a fully automated setup backed by SaltStack running in the AWS cloud.

  • 07/2016 - 11/2016

    • Microsoft Azure; Webtrekk
  • System Administrator / DevOps
  • LANGUAGES Improved an existing DataStax Cassandra Cluster through a multi phased process of collecting
    performance data with prometheus, identifying existing bottlenecks and implementing multiple
    German configuration enhancements, amongst them fixing rack awareness, heap space allocation, number

    English of nodes, data replication level and automation with Ansible. After the changes the amount of
    running transactions per second was increased in around 400%.
    Greek Automated the installation of Apache Kafka brokers by creating custom Debian packages and an

    Ansible role to install them, improving provisioning time for new clusters.

    Deployed a PoC project for Java application tracing based on Dynatrace, enabling development
    teams to proactively detect performance bottlenecks across services and infrastructure

  • 04/2015 - 04/2016

    • The ADEX
  • System Engineer
  • Planned and migrated a Cloudera cluster from major release version 4 to 5, including an linux
    upgrade from CentOS 6 to 7 managed by puppet with all new configurations and components
    (Hadoop HDFS, YARN, HBase, Hive, ZooKeeper, Kafka, Flume, Spark). The existing data was
    synchronized by homegrown scripts with zero downtime.

    Participated in a true DevOps culture in steadily improving YARN. Worked hand in hand with
    development teams to determine the best resources for the various workloads to achieve the best
    possible result, obtaining a performance increase of 50%.

    Creation of various enrichment scripts for 3rd party data consumed from many data sources saving
    on-demand changes to Hadoop HDFS.

    Participation in the construction of a data enrichment pipeline by means of Kafka brokers and
    flumes in the process of sequential data processing, storing data of a defined file size limit in a HDFS
    cluster. The system was able to process 2 million synthetic data messages per second.

    Created a python-based data backup and recovery solution for Cassandra database systems
    emulating 1 automated recovery per month.

    Setup a Gitlab CI 8 build pipeline to support development teams in their build and release process.

  • 09/2014 - 04/2015

    • ad pepper media Group
  • System Administrator
  • Migrated workloads running over bare metal to an OnApp KVM based solution with Samba4 acting
    as a Domain Controller for the environment. Also deployed a unified monitoring and alerting setup
    for this environment with Nagios.

    Implemented a centralized configuration management system with puppet (v3.4) and created
    custom puppet modules for various infrastructure components (MariaDB, Nagios core, vsftpd,
    OpenVPN, shorewall, iptables, HAProxy, POSIX users and groups, PowerMTA, ActiveMQ, Apache,
    MongoDB, sudo, OnApp, NFS, OpenLDAP).

    Provided consultancy on performance tuning for MariaDB Galera clusters to subsidiaries of the ad
    pepper holding.

    Performed migration of Microsoft SQL Server clusters to version 2014 from 2005 with the following
    features: DB-Engine, Analysis Services, Integration Services.

  • 10/2013 - 08/2014

    • UDG United Digital Group
  • System Administrator
  • Administration and governance of the entire IT infrastructure including high availability, security,
    backup on ubuntu linux servers and ruby application and moving software business workloads into
    a software defined datacenter with VMWare vSphere.

    Implemented a unified security solution via multiple Sophos UTM (WAN, LAN, VPN).