Senior Oracle DBA / Engineer

1/2022 – 4/2022

Discovery work pertaining to Exadata Cloud@Customer Database Machines:

Documentation federation
Exadata-related books reading
Analysis of field engineering materials such as internal config- and check-lists
Production of network topology- and hardware-centered schematics (DB and cell server back panels, Cisco RoCE and management switches arrangement...)
Admin-managed to policy-managed clustered database conversion testing
Exawatcher archives fetched consolidation
Scheduled jobs and systemd services inventory
Exadata & Cell Servers-centered database dynamic view querying
...


Having had access to an ExaC@C lab environment, scripted the following in bash (assembled in one convenient support package):

Automated database instance discovery in terms of names, owning Unix groups and logins, supplementary groups, Oracle product home locations ...
Automated .profile file generation on a per database instance basis (i.e. automated discovery of the Unix process environments under which instances run)
Automated O/S platform memory configuration discovery in terms of RAM allocation, usage, and huge pages
Automated Linux kernel tunable configuration gathering for reconciliation against Oracle requirements. (The same applying to so-called shell limits).
Built-up a set of SQL statements conveniently submittable through wrapped SQL*Plus calls generating readable tabbed Excel data (cell servers, cell disks, ASM disks and disk groups, ASM consumers, database instance initialization parameters ...)

Oracle Database, Oracle RAC (Real Application Clusters)

10/2020 – 4/2021

Oracle Database & Clusterware (ASM) 19c on Oracle Linux 7.9 and IBM PowerHA SystemMirror w/ Hyperswap. System architecture, implementation prototyping, technical counselling/ documentation.

Context
Migration of 59 business applications
Agile PM methodology
Technical (non-exhaustive)
Clusterware/ ASM-hosted standalone 19c Oracle database instances
Incrementally-updated backup copies leveraging Fast Recovery Areas
Non-HADR active/ passive HyperSwap-enabled PowerHA SystemMirror stretched clusters
ONTAP 9 and NetApp filers, exposing NTFS security-style file systems via SMB and NFSv3. Implicit user name mapping backed by LDAP-based (Ms AD) user principal provisioning (including Oracle Unix logins).

PM Assistance
Counseled RBC application custodians, IBM project managers as well as delivery teams regarding the following technologies/ methods and their proper implementation:
Oracle Databases
Database data migrations
PowerHA SystemMirror
NetApp's ONTAP
In that regard, liaised/ worked with the following delivery teams/ competencies
Oracle DBAs
Unix engineers
SAN & NAS engineers
Directory & IAM engineers
Established the so-called « Oracle IWS » (Integration Work Sheet), for the purpose of facilitating the collection of requirements / specifications having to drive the provision of Oracle Databases.

Technical Documentation:
Established convention, requirement, as well as specification documentation regarding the provision of Oracle Database services (similarly to what was done for the CJUE)
With regard to specifications, proposed a Bronze/ Silver/ Gold offering

Prototyping & Automation:
Prototyped, validated and documented the platform-agnostic, standardized implementation of The Oracle Clusterware/ Database stack using Ansible 2.9 on Oracle Linux 7.9 running as VirtualBox VMs on 64-bit Slackware Current
Also seized this opportunity to extend platform-agnostic Ansible automation processes to Oracle Linux 7.9 settings (kernel parameters via sysctl, shell limits, huge pages)
This endeavor resulted in the conception of a standardized, near-procedural Ansible playbook framework. Able, for example, to automate the collection and time-stamped naming of compressed log files
Extended to the automation of the following operations:
Database refreshes through RMAN-driven over-the-network duplications
Database backups

Oracle Database, Oracle RAC (Real Application Clusters)

3/2020 – 9/2020

Assistance to State Street with regards to the delivery of an 8-node, two-cluster HPE Simplivity infrastructure platform aimed at substituting for legacy VMware ESXi 5.5-based hypervisors

Such HPE Simplivity infrastructure having had to eventually host a set of 100+ migrated virtual machines, supporting business applications used by the Luxembourg branch of State Street Bank International GmbH

Whose operation fell under IBM GTS Luxembourg's responsibility

Change order abstract:
Coordinate and manage the technical activities of IBM's personnel
Establish and maintain communications through the SSC Project Manager
Assist the SSC Project Manager prepare and maintain the project plan which lists the activities, tasks, assignments, milestones and estimates for performance
Measure, track and evaluate progress against the project plan
Resolve deviations from the project plan with the SSC Project Manager
Conduct regularly scheduled meetings, as necessary, with the SSC project team to review project status
Prepare standard Status Reports through project duration for presentation to SSC Project Focal

Liaised with Ireland- and India-based State Street project management and technical lead, so as to ensure jointly-driven, consensus-based technical decision making as regards OmniStack releases, network and Simplivity topologies

Maintained required communication channels with IBM GTS India, the support of whom got obtained as regards proper engagement of HPE's support from a pure contractual perspective.

Liaised with IBM GTS Luxembourg VMware/ Windows delivery teams involved in the operation/ management/ maintenance of State Street Luxembourg's virtualized environments

Automated the generation of PDF-formatted Project Management Products for metrics publishing, including forecasts and rolling figures aggregated on a per billing code, per competency center basis

Having allowed State Street Project/Program Management to get regularly updated so as to meet contracted Project Management requirements in that regard

PRINCE2

6/2018 – 7/2018

Oracle Database 11gR2 -> 12cR2 Schema Migrations

Within the broader context of the migration of business applications from a legacy network shared by multiple banking entities to an IT infrastructure exclusively-dedicated to BIL, undertook the specific migration of database schemas by porting them from AIX-based 11gR2 Oracle Databases to Exadata X6-based, Linux-hosted Multi-Tenant 12c Containers.

This mainly involved Data Pump exports and imports.
Automated schema migration through shell scripting, encapsulating proper SQL DDL/DML statements
Wrote data dictionary queries so as to perform prior scoping

Schema migrations were required to be automated as much as possible, so as to ease the process for production DBAs. A migration involved a two-step process: first export, then import.

The export and import phases, which could be conducted independently as one piece, could be both run repeatedly until the migration process was definitively established. The creation of the target 12c PDBs was (obviously) a pre-requisite, and occurred once, independently.

Bash scripts encapsulating calls to SQL deliverables handling export and import jobs as well as a base canvas for pre- and post-import SQL scripts, were generated in a unified manner based on a set of input constants such as Unix login names, database instance names, 12c container service names, database account names …

Which, among other things, allowed relevant scripts to reference file system locations hosting Data Pump dumps identically-exposed to both source and target systems via NFS.

Bash shell scripts were written so as to automatically execute, upon invocation, any additional, freely scriptable SQL-only pre-export and post-import deliverables, provided they were prefixed by a two-digit string acting as a simple ordering mechanism, and dropped at the proper file system location.

Oracle Database