Roman Nabilskyi
DevOps Engineer at Windward- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
Topline Score
Bio
William A.
Roman has worn many hats during his time at EZ Texting. Roman was one of the earliest team members in as part of a (then new) NOC engineering team, and was responsible deploying and maintaining EZ's earliest monitoring and alerting tools, settling on Zabbix, ELK, and PagerDuty as the stack for EZ's first-generation observability platform. Logging in particular has been Roman's major focus, and Roman has been central to scaling EZ's internal logging platform as EZ has grown, while keeping costs low. (For reference, as of June 2021, EZ Texting was generating hundreds of millions of log messages per day). Roman was the go-to expert for all things related to operating ELK (EZ's logging platform), and as EZ's observability needs grew beyond the capabilities of the existing tools, he also played a leading role in migrating logging functionality over to Datadog. Roman has also pushed EZ toward more advanced observability functionality. Roman was an early advocate for implementing application performance monitoring (APM) with distributing tracing, and was one of a handful of engineers charged with deploying APM agents (initially New Relic, and then Datadog) throughout our fleet. Today, EZ's observability processes heavily utilize these APM metrics to identify bottlenecks and narrow down problems across over a dozen services. Roman has fought to eliminate toil for the team. In addition to surfacing useless and non-actionable alerts, Roman played a key role in helping to streamline EZ's post-release testing process. This cut the time to perform smoke tests by more than half, which not only made life easier for the engineers involved in the release, but also gave much faster feedback on whether the release was successful, or if it needed to be hotfixed or rolled back. Lastly, when I was tasked with leading a newly-combined NOC and Service Desk team, Roman's assistance was valuable to that team being successful. Roman helped advise me on the team's culture and sentiment, and also helped me socialize some of the efforts I was undertaking to upskill the team and integrate it into a broader SRE team. I am very sad that we are parting ways. If you're looking for an observability engineer (especially one experienced with distributed or cloud-native systems), I can highly recommend Roman.
Marianna Tkacheva
Roman is the best "IT wizard" I've ever worked with: highly professional, utterly tactful & patient in any complicated & intricate situation, always there to help in dire hard-/software need. It goes without saying his considerate behaviour makes him welcomed in any surrounding. He's an extremely bright & "warm" personality with wide scope of interests, thirsty for further advance in all spheres.
William A.
Roman has worn many hats during his time at EZ Texting. Roman was one of the earliest team members in as part of a (then new) NOC engineering team, and was responsible deploying and maintaining EZ's earliest monitoring and alerting tools, settling on Zabbix, ELK, and PagerDuty as the stack for EZ's first-generation observability platform. Logging in particular has been Roman's major focus, and Roman has been central to scaling EZ's internal logging platform as EZ has grown, while keeping costs low. (For reference, as of June 2021, EZ Texting was generating hundreds of millions of log messages per day). Roman was the go-to expert for all things related to operating ELK (EZ's logging platform), and as EZ's observability needs grew beyond the capabilities of the existing tools, he also played a leading role in migrating logging functionality over to Datadog. Roman has also pushed EZ toward more advanced observability functionality. Roman was an early advocate for implementing application performance monitoring (APM) with distributing tracing, and was one of a handful of engineers charged with deploying APM agents (initially New Relic, and then Datadog) throughout our fleet. Today, EZ's observability processes heavily utilize these APM metrics to identify bottlenecks and narrow down problems across over a dozen services. Roman has fought to eliminate toil for the team. In addition to surfacing useless and non-actionable alerts, Roman played a key role in helping to streamline EZ's post-release testing process. This cut the time to perform smoke tests by more than half, which not only made life easier for the engineers involved in the release, but also gave much faster feedback on whether the release was successful, or if it needed to be hotfixed or rolled back. Lastly, when I was tasked with leading a newly-combined NOC and Service Desk team, Roman's assistance was valuable to that team being successful. Roman helped advise me on the team's culture and sentiment, and also helped me socialize some of the efforts I was undertaking to upskill the team and integrate it into a broader SRE team. I am very sad that we are parting ways. If you're looking for an observability engineer (especially one experienced with distributed or cloud-native systems), I can highly recommend Roman.
Marianna Tkacheva
Roman is the best "IT wizard" I've ever worked with: highly professional, utterly tactful & patient in any complicated & intricate situation, always there to help in dire hard-/software need. It goes without saying his considerate behaviour makes him welcomed in any surrounding. He's an extremely bright & "warm" personality with wide scope of interests, thirsty for further advance in all spheres.
William A.
Roman has worn many hats during his time at EZ Texting. Roman was one of the earliest team members in as part of a (then new) NOC engineering team, and was responsible deploying and maintaining EZ's earliest monitoring and alerting tools, settling on Zabbix, ELK, and PagerDuty as the stack for EZ's first-generation observability platform. Logging in particular has been Roman's major focus, and Roman has been central to scaling EZ's internal logging platform as EZ has grown, while keeping costs low. (For reference, as of June 2021, EZ Texting was generating hundreds of millions of log messages per day). Roman was the go-to expert for all things related to operating ELK (EZ's logging platform), and as EZ's observability needs grew beyond the capabilities of the existing tools, he also played a leading role in migrating logging functionality over to Datadog. Roman has also pushed EZ toward more advanced observability functionality. Roman was an early advocate for implementing application performance monitoring (APM) with distributing tracing, and was one of a handful of engineers charged with deploying APM agents (initially New Relic, and then Datadog) throughout our fleet. Today, EZ's observability processes heavily utilize these APM metrics to identify bottlenecks and narrow down problems across over a dozen services. Roman has fought to eliminate toil for the team. In addition to surfacing useless and non-actionable alerts, Roman played a key role in helping to streamline EZ's post-release testing process. This cut the time to perform smoke tests by more than half, which not only made life easier for the engineers involved in the release, but also gave much faster feedback on whether the release was successful, or if it needed to be hotfixed or rolled back. Lastly, when I was tasked with leading a newly-combined NOC and Service Desk team, Roman's assistance was valuable to that team being successful. Roman helped advise me on the team's culture and sentiment, and also helped me socialize some of the efforts I was undertaking to upskill the team and integrate it into a broader SRE team. I am very sad that we are parting ways. If you're looking for an observability engineer (especially one experienced with distributed or cloud-native systems), I can highly recommend Roman.
Marianna Tkacheva
Roman is the best "IT wizard" I've ever worked with: highly professional, utterly tactful & patient in any complicated & intricate situation, always there to help in dire hard-/software need. It goes without saying his considerate behaviour makes him welcomed in any surrounding. He's an extremely bright & "warm" personality with wide scope of interests, thirsty for further advance in all spheres.
William A.
Roman has worn many hats during his time at EZ Texting. Roman was one of the earliest team members in as part of a (then new) NOC engineering team, and was responsible deploying and maintaining EZ's earliest monitoring and alerting tools, settling on Zabbix, ELK, and PagerDuty as the stack for EZ's first-generation observability platform. Logging in particular has been Roman's major focus, and Roman has been central to scaling EZ's internal logging platform as EZ has grown, while keeping costs low. (For reference, as of June 2021, EZ Texting was generating hundreds of millions of log messages per day). Roman was the go-to expert for all things related to operating ELK (EZ's logging platform), and as EZ's observability needs grew beyond the capabilities of the existing tools, he also played a leading role in migrating logging functionality over to Datadog. Roman has also pushed EZ toward more advanced observability functionality. Roman was an early advocate for implementing application performance monitoring (APM) with distributing tracing, and was one of a handful of engineers charged with deploying APM agents (initially New Relic, and then Datadog) throughout our fleet. Today, EZ's observability processes heavily utilize these APM metrics to identify bottlenecks and narrow down problems across over a dozen services. Roman has fought to eliminate toil for the team. In addition to surfacing useless and non-actionable alerts, Roman played a key role in helping to streamline EZ's post-release testing process. This cut the time to perform smoke tests by more than half, which not only made life easier for the engineers involved in the release, but also gave much faster feedback on whether the release was successful, or if it needed to be hotfixed or rolled back. Lastly, when I was tasked with leading a newly-combined NOC and Service Desk team, Roman's assistance was valuable to that team being successful. Roman helped advise me on the team's culture and sentiment, and also helped me socialize some of the efforts I was undertaking to upskill the team and integrate it into a broader SRE team. I am very sad that we are parting ways. If you're looking for an observability engineer (especially one experienced with distributed or cloud-native systems), I can highly recommend Roman.
Marianna Tkacheva
Roman is the best "IT wizard" I've ever worked with: highly professional, utterly tactful & patient in any complicated & intricate situation, always there to help in dire hard-/software need. It goes without saying his considerate behaviour makes him welcomed in any surrounding. He's an extremely bright & "warm" personality with wide scope of interests, thirsty for further advance in all spheres.
Experience
-
Windward
-
United States
-
Manufacturing
-
DevOps Engineer
-
Jul 2021 - Present
Duties: - Deploy services to AWS using Terraform. - Work closely with developers by improving CI/CD process. - Adding new and improve existing Jenkins jobs. - Create new docker images, updating the old ones. - Adding pod templates with different docker images/yaml configs in Jenkins. - Took part in devops escalations and participated in on-call rotations. Accomplishments: - Implemented AWS ec2-instance-connect feature that allows connecting to hosts without persistence ssh key, ability to restrict access to hosts by adding IAM policies for admin/dev/qa. - Upgraded the Jenkins master, slaves moved to EKS, by moving slaves to EKS the queue of waiting jobs was eliminated. - Converted Jenkins jobs from freestyle to pipeline and used shared libraries to eliminate code duplicity among pipelines, created deploy pipeline for java. - Upgraded self-hosted Confluence to the latest LTS. Show less
-
-
-
EZ Texting
-
United States
-
Telecommunications
-
1 - 100 Employee
-
DevOps Engineer
-
Jan 2018 - Jul 2021
Duties: • Production deploys using Ansible and Jenkins, stage deploys in a similar way. • Deploy new VMs in Google Cloud Platform using Terraform. • Push new/update existing Ansible/Terraform configs and scripts into GitHub codebase. • Design and deploy Zabbix|Prometheus monitoring configs for new services, DataDog integrations. • Support and tune ELK stack components to keep up load, special during elections. • Debug production issues using different monitoring solutions - Zabbix, Prometheus, Pingdom, ELK and DataDog. • Run smoke tests after release and hotfix. • Provide assistance for Dev/QA regarding any production/stage issues. Other tools/services that were used: • Scrum / Kanban methodologies. • PagerDuty to catch production alerts. • Familiar with Jenkins, CI/CD. • Bash scripts. • Logstash filters. • Supervisor, Gearman, Redis. Accomplishments: • Deployed ELK stack based on hot-warm architecture together with RabbitMQ as message buffer to handle millions of log messages per day. • Deployed Prometheus and Grafana, setup system/application metrics gathering from datacenter and GCP, setup different dashboards for easier navigation across services. • Implemented different(RabbitMQ, PHP, PHP-FPM, Kubernetes, PagerDuty, Logstash) integrations and setup application performance monitoring for PHP in DataDog, setup dashboards for support team for quick analyse current system load. • Implemented NewRelic integration for Java(application performance monitoring). Show less
-
-
-
CallFire
-
United States
-
Telecommunications
-
1 - 100 Employee
-
NOC Engineer / NOC Team Lead
-
Feb 2015 - Jan 2018
Duties (as noc engineer): • Monitor system stability using Zabbix,ELK, debug production issues, create downtime reports, interact with vendors in case of emergency via phone/email. • Deploy Zabbix scripts on existing hosts(monitoring improvements) and new VMs(monitoring of new services) using Ansible. • Create new and update existing documentation. • On-call rotation engineer. As NOC team lead: • Train, mentor NOC engineers. • Weekly reports to direct manager and brief reports to senior management. • Implement new processes and keep in update existing. • Interacted with other teams to build new monitoring features, new processes. Show less
-
-
-
Innovecs
-
United States
-
IT Services and IT Consulting
-
400 - 500 Employee
-
System Administrator
-
Feb 2013 - Feb 2015
Duties: • System administration / HelpDesk. • Supporting over 150 users in head office and branch, ensuring 24x7 infrastructure performance company. • Evaluation, recommendation, installation and configuration of hardware and software for new projects. • Diagnosis and repair of PC, supply of necessary components (upgrade), the selection of the required configuration of the workstation. • Personnel training. Setup and support: • WinServer 2003/2008r2 • Windows XP/7/8 • Centos 6 • Asterisk • Atlassian products • Office 365 cloud services(Exchange, Sharepoint) • Microsoft Hyper-V 2008r2 • Corporate website • 1C 8.2 (also on Linux platform centos+postgresql) • Backup solutions (clonezilla / norton ghost / wbadmin) Accomplishments: • Setup Hyper-V with different production VMs. • Setup second domain controller for fault tolerance. • Implemented Windows Deployment Service for unattended setup of Windows 7/8 with preinstalled software. • Implemented corporate SSl-VPN based on AD Ldap and Fortigate. • Implemented Zabbix as monitoring solution for availability and performance of IT infrastructure components. Show less
-
-
-
Govitall.com
-
IT Services and IT Consulting
-
100 - 200 Employee
-
junior system administrator
-
Nov 2011 - Feb 2013
Duties: • Technical support using issue tracking system. • Monitoring and installation of existing as well as new workstations. • Connecting and configuring peripherals (printers, scanners, routers, ip-phones). • Installing and configuring of Windows server 2003, AD DS, Group policy. • Configuring ACS(access control system). Duties: • Technical support using issue tracking system. • Monitoring and installation of existing as well as new workstations. • Connecting and configuring peripherals (printers, scanners, routers, ip-phones). • Installing and configuring of Windows server 2003, AD DS, Group policy. • Configuring ACS(access control system).
-
-
-
КНДІСЕ
-
Ukraine
-
Furniture and Home Furnishings Manufacturing
-
1 - 100 Employee
-
any key / help desk
-
Dec 2009 - Feb 2012
First job! Duties: • Helpdesk. • Setup and support different kind of OS's and software, problems troubleshooting. • Hardware support - repairing laptops/PCs and peripherals. • Involved to build structured cable systems. First job! Duties: • Helpdesk. • Setup and support different kind of OS's and software, problems troubleshooting. • Hardware support - repairing laptops/PCs and peripherals. • Involved to build structured cable systems.
-
-
Education
-
2005 - 2010
NUFT (National University of Food Technology )
Master's degree, Information Technology
Community
