top of page


My over-a-decade industry experience have seen me participate and lead many innovative projects, a few of which are listed here:


Deployment platform built to quickly deploy instances with user defined parameters such as operating systems image, computing (memory and processor) capacity, storage etc. upon which the user (developers and testers) can build and test applications.

Requirement Specification

Compute: 270+ vCPUs

Memory: 1.4TB RAM

Storage: 90TB



100+ vms with 1 TB shared and 1 TB private storage

1000+ containers to run workload


Built-in Features and function

  • Multi-node Spectrum Scale deployment using Ansible

  • Multiple Operating Systems Images

  • Multiple Spectrum Scale Versions with –

  1. Multiple NSDs, Filesystems and Storage Pools

  2. Easily Configurable CES feature

  3. Proactive Services Enabled.

  • Spectrum Scale Log analysis with ELK stack

  • Plug and play images (ELK, ML/DL) to avoid hardware and dependent package dependencies

  • Container Workload support (beta)


A 45 node cluster with three controllers serving as the quorum managers. The 6 openstack daemon of Heat, Cinder, nova, keystone, glance, etc ran on the three controllers using the Spectrum Scale cluster files system to enable speedy, concurrent and parallel access to the filesystems of the other nodes in the cluster. This arrangement enables reliability in the sense that when one of the nodes on which the openstack application runs fails, the other nodes in the cluster assume the role of the failing node until it is restored and added back to the cluster.


The result of this project is significant reduction in the time it takes to deploy a node in readiness for testing and development.



An efficient infrastructure organized for developers and testers to run workloads in containerized settings, while operation engineers are tasked with optimal scalability, ensuring and monitoring availability.



1 Bootstrap machine

3 Master Nodes

Per machine

Model: ThinkSystem SR665

Memory: 503GB RAM

Compute: 96

Storage: 12.7T


The bootstrap and master machines were provisioned using Xtreme Cloud Automation Tool. A few of the key advantages of this automated arrangement is that it provides an automated environment that prevents risk of human error during deployment and the containment within a cohesive controller setting.


Once the bootstrap and the master nodes were successfully installed with the appropriate Operating systems, the bootstrap machine was then was then configured to provision the master nodes with the Red hat Core Operating System image and other services which enables them to perform their function of managing worker nodes and scheduling pods.


The result of this project is significant reduction in cost associated with hardware acquisition and speedy pace of application development and testing. Less computing capacity were required to develop and test applications as those applications have now been containerized thereby losing the computing overheads that would have been devoted to them in alternative settings.



This setup was geared towards quick, automated deployment of nodes with varying degree of user intervention as might be necessary. xCat provides a comprehensive platform upon which an entire component of a distributed system can be clustered and automated tools developed to control them.


The result is dynamic and flexible deployment procedure which led to robust application development and testing.

bottom of page