Arghya Banerjee

Objectives :

To evaluate, design, implement and manage large infrastructure Platforms, public

and private ‘clouds’ or ‘Infrastructure as a service’. IT services, application and infrastructure

automation using tools like chef. System and application security.

Skills:

• Openstack

• Software defined networking

• KVM, LXC and Docker on Linux

virtualization

• Automation and configuration

management using Chef

• OpenLDAP (SLAPD)

• Capacity Planning

• System and Process Audits

• MRTG, Cacti, NAGIOS and Munin

monitoring

• IT security

Knowledgebase:

Operating Systems:

Linux (Ubuntu, RHEL, CentOS, Slackware), MS Windows, Solaris and HPUX

High Availability Solutions:

KeepAlived, HAProxy, HP-M/C ServiceGuard HP-UX cluster 11.14 with HP-LVM, Sun

Cluster 3.1, 3.2 with Shared QFS, Windows 2003 Cluster, BigIP F5 / Nortel and

Coyote Network Load balancers / SSL offload engines, Hadoop clusters, CEPH storage

clusters.

Cloud platforms:

OpenStack, AWS, Go-Grid.

Certifications/Trainings:

OpenStack 3-day workshop with Mirantis, Inc.

Chef basic and intermediate training from Opscode / Chef Software. Inc.

HP Certified Unix System Administrator

BigIP F5 LTM 1600 basic training

VMWare PowerCLI training

• AWS

• Software defined storage

• CEPH Distributed and highly

scalable storage systems for private

clouds.

• OpenVswitch

• BigIP F5, HAProxy Load Balancers

• Ubuntu/RHEL/CentOS LINUX

• OpenLDAP (SLAPD)

• Cloudera Hadoop Clusters

• VMware ESXi, Amazon and GoGrid

Cloud

Databases:

Percona XtraDB MySQL clustering and PostgreSQL on Linux, SQL Server 6.5/ 7.0. /

2000 on windows NT / 2000 / 2003, ORACLE on HP-UX 11i, Solaris 9 / 10, RHEL 5.1 (x64)

and Windows NT / 2000 / 2003,

Languages:

C, C++, Pascal, FORTRAN, BASIC.

Development tools:

Visual C++ 4.0/5.0, MFC 4.0, Windows SDK, Borland C / C++ 3.0., Ruby, BASH and

PowerShell scripting.

Firewalls:

IPTables, CheckPoint, ConSEAL & Tiny personal firewall.

Experience:

Sr. System Engineer - IT: Adobe Systems, Inc. San Jose, CA, USA.

From September 2012 till date.

Adobe Systems, Inc. provides creative suite of applications as a service for web design,

graphics desgin, document management and analytics suite of applications as a service for

sales and marketing organizations.

Projects at Adobe IT:

• Automation using Chef – Set up multiple chef servers in different engineering and IT

environments to provide chef server as a service with high SLA for internal Adobe

groups to consume. Evangelized application configuration and management automation

using chef cookbooks by closely working with project admins and technical architects,

helping write the cookbooks. Hadoop and OpenStack infrastructure deployments using

chef.

• IT IAAS – managed Linux and Unix servers as part of the Unix platform team in IT.

Helped automate manual processes and regular maintenance activities using VMWare

PowerCLI scripts, chef cookbooks and Bladelogic jobs.

• Hadoop – helped analytics team set up a 60 node Cloudera Hadoop cluster. Automated

the deployment, tuning and management of linux servers used for the hadoop cluster in

a seamless and quick manner using chef cookbooks.

• Openstack – Designed and deployed the first folsom version of OpenStack cluster as a

technology demonstrator in IT (2013). Evangelized OpenStack and its advantages using

the Folsom cluster at internal conferences and architecture summits. As part of the

OpenStack architects team, I designed and deployed IT’s first “production” OpenStack

cluster using IcceHouse OpenStack on Ubuntu (2014). Deployed -

o “vLAN provider network”,

o CEPH based 300TB size distributed storage cluster (Deployed FireFly version

and now upgraded to Hammer) with separate storage network and CEPH

storage for VM workloads,

o OVS software switches with OpenStack Neutron ML2 plugin,

o linux bonded interfaces on the bare metal hosts in OpenStack and

o vxLAN overlay networks for “tenant private networks”

o Chef cookbook based deployment automation for critical OpenStack components

o Deployed stable and tested workarounds to OpenStack IceHouse bugs/issues,

backported code changes and patches from mainstream OpenStack Juno release

to our IceHouse version.

o Load and functionality tested OpenStack services and API (provided proven

stability up to 60 simultaneous instance spin-ups and 5000 simultaneous MySQL

connections on the backend)

• AWS – Worked with IT’s cloud architecture team in deploying IT applications in a secure

and supported manner on AWS cloud. Set up Chef server in AWS for configuration and

management of IT applications in AWS.

• Vagrant, VirtualBox and Chef – Worked with engineers and software architects in

designing application lab, development and staging environments for them using

vagrant, Vitualbox and chef on their individual laptops or desktops. Helped create 'IT

Certified' chef cookbooks for automated deployments of mongodb and Percona Xtradb

MySQL clusters.

Operations Manager / Sr. System Administrator: Tavant Technologies, Inc Santa Clara, CA, USA.

From January 2008 till August 2102.

Tavant Technologies, Inc. was an e-business solutions provider for large and mid-size

manufacturing companies, banking institutions, and video game companies. They provide

system design, development and SaaS hosting solutions.

Projects worked on at Tavant as the Team lead and currently as Associate Manager for all

system administration operations in the SAAS group:

• Project Manager and Lead System Admin – Hosting migration of all Bobcat

North America Internet Sites/Services – Managed the migration of all Doosan

/Bobcat North America Public Internet services to Tavant hosting between April 2011

and August 2011. The migration was planned and implemented flawlessly between

Tavant and Doosan North America teams. Internet facing bobcat.com, bobcatnet.com

and other public services for Bobcat dealers were migrated. Planning included adding

capacity to our hosting data centers, internet bandwidth at the hosting sites, upgrading

database servers, storages, and application servers. The migration also involved

migrating “physical servers” to “virtual machines” on VMWare. Application performance

on all migrated applications improved as we migrated from older architecture and

hardware to newer servers and newer architecture with network and database servers.

• Storage Array upgrade – We upgraded the Storage Array used by our Staging Environment

Oracle RAC cluster to a HDS (Hitachi) array in February 2011. I worked with vendor for hardware

install and then created the LUNs and presented them to the hosts. The LUNs were carved out to

facilitate the RAC cluster with Voting disks as well as data storage disks in different

configurations.

• Hosting Solutions remote control panel – In 2010 I created a control panel writing

a library of Unix Shell CGI scripts to remotely stop/start/restart production services

and also extract java thread dumps from the java services. The control panel is

password protected and exposed on a non-standard TCP port over SSL. I have also

created a more intrusive , Menu-driven Unix shell script to manage these same

production services over SSH and company VPN. This shell menu can be used to do

more in-depth trouble shooting (do a code release if required). The Web based control

panel is tested and works great with certain Smart-phones (which was my goal in

creating the control panel).

2008 to 2010:

• Under my guidance the operations team architected and implemented the entire Oracle

database server hardware architecture with RAW volumes (oracle ASM) for Oracle 10g

RAC running on 64-bit RHEL 5.1.

• Designed and created the storage SAN layout in EMC Clarion CX3-10 series arrays. We

used Oracle ORION I/O tool to benchmark I/O performance for target databases on

the servers using database characteristics from older environment.

• We migrated all application servers running Oracle/Bea Weblogic, Redhat Jboss and MS

IIS servers from older HP P-Class blades to newer HP C-Class blades.

• I tuned RHEL 5.1 x64 for Oracle 10g RAC, Oracle OCFS file system and ASM raw file

systems.

• Created and designed hardware, network layout to support multiple Oracle RAC clusters,

up to 6 EMC Arrays and Fiber SAN switches.

• Consolidated four data centers into two: Migrated about 70 application servers,

Database servers, DNS servers, Sendmail servers and network equipment from four

data centers into two new redundant hosting data centers.

• We deployed redundant BigIP F5 1600 LTM Load Balancers.

• Deployed applications in the cloud environment as well as VMWare virtual

environments.

• My team provides day to day operations support, code release support and new

customer roll-outs to our Software as Service operations service. Currently Tavant’s

warranty management systems process about 1000 warranty claims a day.

• My team along with Network Admin and DBA teams in providing security and audit

support (SAS70 compliance audits) and business continuity audits for customers.

• I have helped create processes for releasing and maintaining code to our hosted

applications.

• I am constantly providing guidance to my team, conducting appraisals twice a year,

recommending training and skill development ideas, keeping management informed

about team activities and project work and progress. I have create Unix script tools for

the team to help provide faster response times and better SLAs for our hosted

customers.

• I also stand in, when required to, as a backup for my supervisor (the Director of

Operations) keeping the hosting (SAAS) group performing smoothly.

Team Management – Leading the team of system administrators –

• Providing guidance,

• Conducting appraisals

• Recommending training and skill development ideas to team members

• Keeping management informed about team activities and project work and progress

• Providing backup support to my supervisor in business development and other

management activities.

Lead System / Network Administrator: Client Network Services, Inc. Gaithesburg, MD and

Augusta, Maine USA.

From October 2001 to December 2007.

Client Network Services, Inc. (or CNSI) is a system integration company for state and

federal governments based out of Rockville, MD.

Projects based on Sun Solaris / HP-UX and Sun, EMC and HP hardware:

State of MI MMIS – Lead system administrator and architect for designing and

implementing the entire solution which includes Sun Cluster 3.1 with RAW volumes for

Oracle 10g RAC. Supporting Webmasters managing the WebSphere App server 6.1.

Supporting similar solution as State of ME MMIS – only larger in scale and on Sun H/W.

State of WA MMIS – Lead system administrator and architect for designing and

implementing Sun Cluster 3.1 with Shared QFS supporting Oracle RAC 10g. Tuning and

hardening Solaris 10 for Oracle 10g RAC database of 8 TB in size, WebSphere App servers

6.0 and 6.1. Supporting similar solution as State of ME MMIS – only larger in scale and on

Sun H/W.

State of ME MMIS - Till November 2006 - was the resident system / network administrator

for Maine State Claims Management system (or MECMS) at Augusta, Maine. I have just

handed over responsibilities to the state of Maine staff and have moved to corporate office

as the lead on infrastructure deployments for all projects for the company as mentioned

above. The computer systems being used for the Maine project were HP 9000 series super

scalable Unix servers and server clusters. I was supporting other modules by administration

of Business Intelligence software - COGNOS, Oracle Financials Applications and

WEBLOGIC Application servers. Worked with program developers and DBAs in rolling out

the application module-wise in different stages. Gathered and analyzed the project

requirements from the client, prepared analysis and design documents and was a major

contributor during the implementation phases. Was responsible for configuring the computer

systems and providing any network and system related support to the program developers

as well as to the client. Monitored and modified the ServiceGuard Cluster which is running

the Primary Database. Created and managed large 4 to 5 TB Unix VxFS (Veritas Online

JFS / LVMs) volumes housed in a XP1024 Disk Array for all the servers. Managed

Security and patch updates on the servers.

Worked with an Oracle Financials DBA, Developer and Functional person to roll out the

Financials section of the Claims Management System.

The application servers were BEA Weblogic servers, load balanced using Nortel’s AD3

switches. The application was HTTPS based with SSL encryption provided by Nortel’s iSD

SSL off-load boxes (which were also load balanced by Nortel’s AD3 switches).Installed the

backup environment with HP OpenView DataProtector 5.5 and HP MSL5000 series tape

library. As part of my responsibilities I managed a staff of one mid-level System

Administrator and also interacted / co-ordinated with 8 State System / network admin.

Wrote complicated shell scripts (using awk and sed) to implement functionality which

was crucial for the project automation.

After the project went live I was instrumental in working with Java Application developers

and Oracle DBAs in performance tuning the large (2 TB) Oracle OLTP and OLAP

databases and Weblogic Java application servers. Was responsible for the CNSI, Augusta

office systems, network and communication circuits. I did the set up of all computers,

servers (file, Licensing and email), Network (WAN and LAN), phone system (which included

ISDN lines for video conference and a voice messaging system integrated with the Exchange

email system -CISCO UNITY), network printers and other parts of the office infrastructure. I

had also set up a broadband (cable) and Wireless Ethernet Network (CISCO WLAN) at the

corporate residential guest house. Maintained and implemented any local network and

system requirements in the CNSI office.

Made part of the Technical Management Group within CNSI in Nov-2006 as Sr.

Infrastructure specialist to consult on CNSI's other two state Medicaid Projects as

mentioned above. As part of my current responsibilities I am managing four mid-level unix

system administrators and guiding them on the new projects based on lessons learned from

Maine.

Projects based on Microsoft platform and Dell / EMC hardware:

State of GA Vital Records – Lead system administrator and architect for designing and

implementing a MS windows cluster. SQL 2000 cluster is running on this MS cluster.

The SQL database is running off a Dell /EMC disk array and is about 300 GB in size.

Tuning windows for the database was part of my SA team. The project also involved tuning

MS windows 2003 servers for Sun One Java Application server and .NET IIS servers. There

are 4 Sun one Java front end Web servers sitting behind a Coyote Web Load Balancer and

SSL offload engine and 4 IIS servers again load balanced by the Coyote load balancer.

Consultant System Administrator: NeedBrains Inc. San Jose, CA, USA. (Employment with eCalyx,

Inc. CA, USA)

From April 2001 to August 2001.

· NeedBrains Inc. was a consulting company based out of Milpitas, CA and San Jose, CA. In

my position as System / Network Administrator I was responsible for setting up the

company’s Unix (Solaris on Sun SPARC, Linux) infrastructure that provides Web

(iPlanet), DNS (secondary), and Email Forwarding (Sendmail on Linux) services. The

company’s main Network (User network / LAN), also completely built by me is based on

MS Windows NT with services like Email (MS Exchange Server 5.5), DNS (Primary),

Proxy (Analog Proxy Server) and Weblogic, JRUN for the development environment.

The needs of the sales, accounting and management staff have dictated a groupware

type environment for centralized storage and easy sharing of data, and hence the

Exchange server for mail (as well as scheduling, contacts, and calendar). The Mail

forwarding server on Linux is to save money and also provide a minimal level of support

for an old company (essentially emails) that existed prior to formation of Needbrains.

The Analog Proxy Server (for sharing a T1 line) and also Tiny Firewall Server

(protecting the Exchange Email and development servers) are both free and also hint at

how far the Dollar can be stretched. However, both the proxy server and the firewall are

very good in doing what they are supposed to do (of course correctly configured by me).

All servers are secured up using Firewall or YASSP and SSH. Also I have successfully

set up and configured a LDAP server using OpenLDAP on Linux for displaying

organization charts and customer info. For me the size or amount of capital influx in the

company is no hindrance for setting up and providing a strong and effective

infrastructure.

Consultant Unix Administrator: VelocityHSI Inc. Walnut Creek, CA, USA. (ISP – Core Group)

(Employment with eCalyx, Inc. CA, USA)

From August 2000 to March 2001.

· VelocityHSI was a provider of high-speed Internet connectivity in the Multi Tenant Unit

environment. In my position as Consultant UNIX Administrator I was responsible for

setting up the company’s Unix (Solaris) infrastructure that provides Email (Sendmail),

News (NNTP), DNS, Billing and Web space (Apache) services. My team is responsible

for the four Sun servers that host the various services in the domain. The email and

Web space for all velocityHSI customers is handled by a Sun E4500 server hooked up

with two redundant StorEdge A5200 disk arrays over optic fibre. The E4500 also acts

as a secondary DNS for our domains. A Sun 220 R rack server carries news and

Primary DNS services. Company web site and user Web Sites sit on a Netra T1 server

and another Netra T1 acts as the firewall protecting our servers. This entire architecture

was designed, set-up and now managed by my team consisting two (myself +1)

engineers. Securing the Solaris 2.7 running on the boxes by cutting down on the number

of daemons running on them with the help of a program called “YASSP” was also done

by me. In order to provide a good and disruption free service I also installed a number

of programs on the unix boxes like SSH (for secure remote administration), tcsh and

the environment that goes with it (for us administrators, programmers and also our

subscribers), BIND 8.2 for DNS and traffic monitoring on the routers with MRTG

running on a Compaq Linux box. Also setting up the Billing environment for the ISP

was one of my tasks, which included setting up of a JRUN server, Weblogic Server and

Apache (ON UNIX) and RODOPI (on NT). I am also part of the core technology group

at the ISP where all technological paths and future options are discussed and then plans

made.

Network / System Administrator : Matrix Information Services Ltd. Bombay, India.

From March `99 to June 2000.

· Matrix (a Kotak Mahindra Group company) was a pioneer in E-commerce in India, selling

copyrighted information over the Internet. In my position as Network / System

Administrator I was responsible for the management of the company’s office setup as

well as the WAN spread over three locations in Mumbai (Bombay). Managing the

company WAN connectivity has enabled me to understand ISDN technology (ISDN is

the main-stay of the company’s WAN connectivity). The network at the company’s hub of

operations is divided into two subnets (10 Mbps and 100 Mbps) with the Windows NT PDC

acting as the peering point. I handled configuration and smooth operation of IP routing,

DNS server and Email server, running on the PDC.

There are three Sun Servers. The live Web Server running Netscape Enterprise Server

and live Database server running Oracle are two separate Sun servers. (This set-up was

a complete suite of products from Netscape called the Netscape Publishing System).

The third Sun server acts as the backup/development server running both the Web and

Oracle Servers. Maintenance and monitoring of the SUN servers’ health on a daily basis is

taken care by me. The main/live servers and the internal network were protected by a

firewall (CheckPoint). Monitoring of Firewall logs for intrusion attempts and unwanted

network activity were also part of my responsibility.

I primarily handled the Year 2000 compliancy project at the company – This included all

the desktop computers, Servers, Telephone systems and also collecting Year 2000

Compliancy certificates from other utility companies providing services to us.

A security audit was done by KPMG of the entire IT infrastructure at Matrix.

Implementation / adhering to most of the recommendations which include securing the

servers individually and the company’s complete IT infrastructure on the whole, were

completed by me (Details are below). I was working on a standardisation program to

streamline and automate (wherever possible) the working of the systems group

(Software professionals group) in the company.

Co-ordinating with infrastructure providers of WAN links, service providers for equipment

and vendor management were also handled by me.

System Security Audit Experience (Security Implementation at Matrix, Bombay -

http://www.indiainformer.com/) (March 1999 – June 2000):

Matrix was an On-line Information Services company in India. I was responsible for setting up a

Security and Responsible Usage policy there with recommendations from KPMG, India. Setting

up a policy on how data is thrown by the web servers, how customers log-in into the servers and

also access control is implemented (both at user level and at Firewall level), what are the

procedures for handling raw and formatted(classified) data by employees within the companies,

what are the access right for employees (depending on their responsibilities eg – Webmaster,

Programmer, Marketing), how raw data and formatted data is backed up and stored securely

(at an offsite underground Bank locker). My responsibilities were to pen these policies / procedures

and also make sure that they are followed by the respective persons (eg – Webmaster, DBA,

Programmer, Classification Dept. etc). Email policy, FTP access policy, remote accesses to the office

network were also checked by KPMG.

Network / System Administrator : Tri-angle Simulation Pvt. Ltd. Bombay, India.

From December `97 to February ‘99

· In my position as System Administrator I was responsible for the management of the

company LANs spread over two locations. I was responsible in implementing a TCP/IP

and windows (NetBEUI) link between the two remote locations. The requirements from

me varied from maintaining the network hardware (to a certain extent) to maintaining

the complete software installed base. Installation and configuration of speciality

softwares like AutoCAD, visual programming tools, etc. and training new users on their

use were taken care of by me. Maintaining a sustainable Uptime of the constantly

changing computer and network environment were a constant challenge for me

(considering the limited resources of the company) and which, I believe I handled

almost perfectly. I had two technicians (deputed from time-time from the

instrumentation dept.) helping me out on the cabling and computer hardware

maintenance. A primary chunk of the task performed by me daily, apart from the ones

mentioned above, was to maintain the NT server, SCO server and the modem link

(scripted dial-up) between the two offices. The systems handled by me were MSWindows

3.11, MS-Windows’9x, IBM OS/2 Warp 4.0 as desktop clients and SCO UNIX

ver 5.0 (OpenServer), MS-Windows NT 4.0 as the two servers connected over TCP/IP

and NetBEUI.

Systems Analyst : P B C International. Bombay, India.

From September `96 to November `97

· Setting up of an Intranet using the company's installed LAN were one of the first tasks

taken up by me. For the Intranet, Microsoft Windows NT 4.0 Server running

Internet Information Sever 3.0 with Microsoft SQL server 6.5 as the database

server and Linux 2.0.0 running Apache ver 1.2 beta are being used for various needs

of the company. Also involved in integrating the various Linux and Windows systems

using Microimages PC X server. I was also responsible for developing and managing

Internet Web Sites on SUN SOLARIS (remotely) for the company and its clients.

Designing of Shopping Carts, Forms and counters on Perl for the Web Sites.

Winsock Project: Vijay Mukhi’s Computer Institute

From September `96 to November`96

· Writing an application in Visual C++ 4.0 using Microsoft’s Winsock ver 1.1 that

registers a company's or individual’s URL at 15 Search Engines on the Internet (for that

time). The project with its entire code is available on the Internet at

http://www.vijaymukhi.com/vmis/addurl.htm.

Trainee - Intern: Vashi Chemi-Equip Pvt. Ltd.

From August `95 to August`96

· Control and administer a small LAN of 5 computers running DOS and a Dot-Matrix printer

connected using a Co-axial cable network. My time there helped me learn the essential

qualities required to work in a Corporate environment, work on some basic computer

and networking skills which proved very helpful in the future.

Education :

University of Bombay, Bombay (Mumbai), India

B.Sc. in Mathematics -1995. (Bachelor of Science)

Vijay Mukhi’s Computer Institute, Bombay

Advanced course in C, C++, Windows programming and Internet Technologies - 1996.

Personal :

Reading novels (fiction, thrillers) & magazines (science, computers and automobiles).

Collecting stamps, pens and stickers.