I have gone from playing with kubernetes to being someone who manages and improves 4 eks clusters under the guidance of someone who has over 8 actual years of kubernetes experience.Lots of terraform to maintain the clusters as code. Not using the console where possible.Making sure all AMIs, dockers and deployed pods are updated every 60 days. Mostly using helm charts, packer, and terraform.Responsible for RDS maintenance. Upgrades and tuning.Took opensearch from a non performing cluster to a well tuned and monitored cluster. Had to do many tweaks and forced data moves, index policies, shard sizing to get it running in a healthy state.I work a lot with Compliance to make sure we maintain a secure environment.

Most of my work was execution of strategic projects. As a subject matter expert with significant experience in engineering, system administration and DevOps, I was brought into many conversations where people just wanted general advice and mentoring. I did not have many standard daily tasks.While it wasn’t a formal part of my role, I took the initiative to mentor junior SRE and DevOps engineers. I feel it is very important to help develop the future generation.A significant amount of my time was spend ensuring a strong cloud security posture. Whether it was IAM, network security, container security, data protection and encryption, protecting our data, and our customers data was a priority for me. I am always challenging the status quo and looking for opportunities to improve our posture. Of course, it is also important to find the balance between security and functionality.Some of the strategic projects I worked on was:- Migration from a department GitHub org to the company Enterprise GitHub org.- Migrated from classic Jenkins 1.x to Jenkins 2.x. Moving from freeform to jenkins file is a major improvement in managing and tracking pipeline changes.- Migrating Secrets management to AWS Secrets. It streamlined the process to ensure adequate protection and properly documented the end-to-end lifecycle of secrets management- Migration from AWS users to OneLogin Roles to Azure Roles.- Python 2 to 3 and java 7 to 8 to 11 to make sure the platform was running on maintained software versions.- To ensure a level of automation and reproducibility, I took the initiative to learn Terraform which I have leveraged in our existing accounts.- As a result of the level of cloud maturity, Lattice systems were typically selected for POCs for many of our security tools.

The first years we had a traditional datacenter setup using VMware. When I started there was a half dead puppet deployment. I replaced it all with Chef. Then we moved to AWS. We decided that the micro services were deployed using cloudformation and ECS. The support systems, jenkins, zookeeper, consul, nexus and others were deployed using opsworks. As is usual for jobs I take and enjoy, I walk into a lot of technical debt that needs to be cleaned up. I also self manage. I send out a plan at the start of each 6 week cycle to all the engineering managers so we can plan my tasks. Mentoring people is also part job. Junior engineers, engineers moving from windows development to linux, customer support engineers, data scientists, any people who are willing to learn. Documenting the crap out of everything is also a big part of the job.Some of my accomplishments:- Migrated services from ESX to AWS. - Migrated from jenkins to latest Jenkins using pipeline and Jenkins DSL. - Migrated from SVN to github. - Migrated from splunk enterprise 6 to splunk cloud 7. - Used ansible for deploying test systems. - AWS Linux to AWS Linux2. - Wrote some code and documented a process to have people deploy code to lambda instead of use the inline editor.- Migrating from Python2 to Python3. Managing the company wide effort, not just my code.- Create a standardized Lambda workflow.- Assist with the ISO 27001 annual security audit.- security operations

Worked on the initial rollout of Chef. Allowed us to not depend on AWS for all our systems. Managing about 500 systems using chef. Maintaining the rolling deploy code for game deployment on 3 games. Working with both AWS and baremetal host provisioning using libvirt/KVM. Coded in python, ruby (chef), BOTO and bash. AWS services used: EC2, cloudformation, autoscale, Eleastic Beanstalk, Route53 and S3. Bare Metal is virtualized using python script wrapping libvirt called from chef recipe. My main focus in this has been dealing with technical debt. The person before me wrote his own system instead of using cfengine, chef or puppet. Reworked various pieces of the game deploy, cloudformation and other internal scripts that hardwired various aspects.I am responsible for the python code used to deploy the 3 games. I set up rolling deploy for one game and a basic deployment scheme for a other. Improved the rolling depoy for the 3rd.Responsible for maintaining the existing Jenkins server and the existing Ganglia setup. Brought up 2 new Jenkins servers online as well as icinga/nagios.I like using chef even if for a single system. Then it fully documents how to replace the system if it is ever needed.On the OPS side, worked with the Game developer teams when issues arose over game deployment. I am responsible for managing the game servers. Have pager duty couple times a week and am on call. I work with engineers in the US and folks in multiple timezones.I balance the daily fires in a live system with improving the automation so the fires are fewer and not as big.

Build & Release Engineer for Rambus. When I started there was no process, no automation, no traceability. Builds came off of a build server or people desktops. SVN, Ant, Puppet and cruise control was used.Currently we use Jenkins, and all builds are automated. The builds are Java, C, Linux, iOS and Android. Some 3rd party opensource software is repackaged. SVN, Ant and puppet are still heavily used. Git is being added to the mix. Automation scripts are written using python, groovy, bash and ruby.Build machines are defined in puppet. 3rd party tools that were being built within puppet are properly packaged into RPMs. RPMs have a private YUM repository. For the one maven build we have a local cache using Archiva. Using Atlassian tools for infrastructure information. JIRA and Confluence is hosted using Atlassian on demand. Fisheye/Crucible is hosted locally.All infrastructure is maintained on Rackspace cloud instances. SSL and authentication is required for all access. Using apache, jetty and tomcat.For the 3rd party builds, RPMs, automake and autoconf are used. Truthfully, creating the configure script using automake and autoconf is something I've been wanting to learn for a very long time.I am not just running Jenkins. I've written Groovy scripts that manipulate the jobs directory and the config.xml directly.I work with Engineers in Milan, San Diego and Sunnyvale. I help a smaller group in Rambus with their SVN. I work with Management, QA and engineering on build and deployment issues. Help coordinate deployments between QA, stage and production environments.The QA team is really a group of expert users and I advise them on basic process.

Build & Release - – Automation Tool Developer August 2011-August 2012- Responsible for all branch requests and 3rdparty tool integration.- Written tools for Parabuild, Confluence, Jira, Crucible, Fisheye in python using both restfull and xmlrpc.- Wrote a tool that added new build to parabuild for the branch requests. Save much time and eliminated many errors.- Wrote a tool that aged out old Crucible reviews and sent waning emails.- Wrote a CLI tool that allowed Cruible review creation.- Fixed, documents and cleanup the test build system. Code in multiple places. Results page hard to read. Notification emails not complete.- With both process, documentation and some code change, the Game certification group has an easier time collecting test data.System Integration Group – Automation Tool Developer May 2008–August 2011- Client automation tool. Setup a client and run a script to test the service. This tool required interfacing with multiple services to get a client, create an account, reserve a apphost, get the current list of games. QA could run games on the service. For OPS could test all the hosts in a newly installed site.- Test Appliance. Wrote a test bed for one of our boards. Loaded latest software, booted board and ran expected tests.Meant to be run on factory floor. Was used for testing the initial boards. Hundreds of units tested. Load Test. This one was two attempts. Learned about the Python global lock the hard way. Second attempt was interesting. Wrote python script to create the client installer for each testhost and the unique information for each simulated client. Recompiled screen to talk to 1000 sessions. Ran a screen session to ssh connections that ran more scripted screen sessions for each of the 80-100 clients on each machine. Screen of screen sessions. It was more fun writing the instruction on how to control screen to the QA guys. Controlled over 1000 simulated clients

I Work within the QA group writing tools to help automate the QA process. Using Perl, autoit, VMware and C# to write a tools that runs applications that capture performance data into a DB and emails a report of the run. This is being used for both performance testing and smoke test automation.Working with Engineering to improve the way unit tests are defined and run. Flush out issues of both how well the tests cover the existing code paths. Also flush out how well assumed functionality is coded. This will require documenting the existing SDK to create more detailed SDK documentation..Working on a C++ test framework that lets the QA engineers define testcases and testsuites in XML. The design allows for the QA engineers to use nouns and verbs they currently use to describe the test cases in XML files that describes a test case. One of the primary goals is for the QA engineers to be comfortable defining the testcases in XML when they are not software engineers.

Build and Tools guy. Make the life of the R&D engineers easier.Created an automated smoke test process that makes use of VMware, expect scripts, automates it to create a multiple VMware VM test environment using the installers from each nights build.Automated the nightly build. The build occurs over 4 VMs. Was a manual process.Migrated from CVS to subversion and reorganized the source tree into discrete components. Currently we build components as a side effect of a monolithic build.Mostly used Perl, expect and bash scripts.

Collected and presented metrics and performance statistics for Yahoo Sites. Most of this first year was spent improving an existing application, Automating processes, Fixed many bugs and Normalized the DB so that query response was acceptable. Also added a series of features. Spent a lot of time learning the Yahoo specific technologies. The system a web based and was written in Perl, Javascript.In 2006, the application was rewritten to take advantage of the knowledge that has been gained about the system users. This was my first foray into the SCRUM style of Agile development process. It was writtenusing PHP, Perl, Javascript and MySQL.

Have been working CM tool and process for this company. We now have all groups using CVS. We had 3 CM products in house. We have standardized on ANT and PERL where possible. The MS Client still require some batch scripts.Have rolled out CVS, CVSMonitor, ViewCVS, Bonsai.Have developed a sopisticated ACL scheme for CVS. Currently working on a web front end.

See Zone Labs, A Check Point CompanyThis entry was just how someone connected to me. See zone labs.

Worked half time on development projects and system administration contracts.I worked on a Java application (JSP and Javascript) application for a couple of Dotcom startups. Spent time at KLA doing embedded realtime application development for a prototype product.Did some XML schema and code development for another dotcom startup.Worked with staff at Sandisk to develop a test environment for one of their chip lines. Created a custom RedHat distibution that required it run on a 100MB flash drive. This distribution was pushed out to over 8000 test stations worldwide.

Senior developer on the VisiBroker Manager 1.0 team. Manager is a tool that presents GUI interface for CORBA’s Interface Repository, Naming Service, Implementation Repository and object query service. The Naming service browser was able to make management of the Naming service feel like using the Windows 95/NT Explorer. Used Java and Swing.Wrote the GUI wrapper for the VisiBroker GateKeeper. Written using Java and Swing.Developed a database report writer written entirely in JAVA. Performed WYSIWYG editing via a preview window. Capabilities included were: nested break-by/Grouping, group/grand totals, font control, text alignment, report save/restore, report page Layout and JDBC interface.