Aleksandar Zivanovic

Account Migration Tool (AMT) Project (Core Java): I was a developer on an AMT ETL tool used to transform database schemas and migrate all the client data to the newer versions of the product. It was an ETL tool with the addition of forking the destination to various product flavors (virtual hosted, clustered, and on-premise). We deployed several successful releases over nine months of the contract.

Adobe Connect Managed Services (ACMC) Project: I was a DevOps developer on a Houston controller written in Ruby, working to reflect rapid changes on the Amazon AWS API side. The controller was used to provision Adobe Connect to AWS cloud. It was similar to how Terraform is used today, with a proprietary source designed only for the Adobe Connect AWS infrastructure. I maintained Cygwin shell scripts and batch scripts inside AMIs and wrote new scripts for operating system actions such as upgrades and syncs. I migrated customer environments to new clusters by deploying rolling release strategies.

Adobe Connect Project: as a software quality engineer for 18 months, I wrote and executed manual test cases and wrote Java-based automation tests for the new application features, using a Jenkins CI setup.

I was a member of the six-person DevOps team that focused primarily on infrastructure and build processes, CI/CD pipelines, and CI/CD improvements in every sprint with features like build plans, deploy plans, testing pyramids, gates, Veracode, SonarQube scans, JaCoCo code coverage, and AppDynamics code.

In addition, we were tasked with server architecture and code deployment automation, maintenance, monitoring, virtualization, and orchestration, using the latest technologies like Docker, Swarm, and Rancher. We managed the basic Atlassian sysadmin product line, including Bamboo CI server, Jira, Stash, Confluence, and Crowd, as well as administration of the virtual machine infrastructure over VMware vSphere in multiple data centers in three locations.

We refactored infrastructure from VM code-based to Docker-container orchestrated and severed configurations automation with Ansible. We performed classic Linux server administration and troubleshooting and optimized environments for the NextGen application. Finally, we integrated build plans with deploy procedures for continuous delivery pipelines for various customers.

I was a TechOps and DevOps engineer in a five-person team called international operations and reliability (IOPS-REPORTS). The team covers five products, mostly in the reporting sector, that process a large number of records, store them in various database engines, and perform transformations through BI tools. All the teams we communicated with daily were T-shaped with full-stack engineers.

The scope of technologies used was vast:
• Old projects ran on bare-metal servers with Sybase and an old, in-house Perl configuration management system called Rubix
• Bash scripting over systems that ran virtualized on Citrix XenServer pools and used Oracle databases, NetApp shared storage, and Citrix NetScaler software and hardware load balancers with various software on the systems, from Python over Java to .NET
• The newest public and private cloud projects on CloudStack and Amazon AWS with Docker containers and Chef and Ansible configuration management, built with GitLab CI and deployed to AWS with Terraform

We refactored one system at a time, making it IaC with CloudStack, Docker, and Chef technologies for the internal cloud and Terraform and Docker for the external cloud projects.

This is a Python Flask-based web server in a uWSGI container that fetches Amazon AWS IP ranges and exports them in the server's root as a text file. Since Palo Alto firewalls can fetch HTTP resources and resources from the local file system, this was one of the solutions to refresh Palo Alto ranges.

I improved the Hewlett Packard Enterprise (HPE) iLO exporter that monitors various parameters of HP Blade hardware in enterprise data centers. It's instrumental in integrating these metrics with other metrics in your Grafana set.