Hi,
Please go through the following requirement and let me know on Dhruv@riderconsultinginc.com if you have any good resources
Role : Data Architect
Duration : 3-6 months
Location : Pleasanton, CA or Chicago, IL
Start: ASAP
Interview : phone
Key Skills: Data Architecture (Modeling, Design), Oracle, Partitioning, Archive/Purge patterns
Background:
· OTP is a legacy OLTP application with about 5Tb of data spanning across 1400 tables
· It is an ASP/Cloud model, so we have about 250 clients on the system
· Data access is primarily via embedded SQL in the application
· Archiving has never been put in place, causing a constraint on online requests
· Our conceptual architecture is to pull old data off the system and place it in a duplicate read-only version of the system for archive retrieval
· We plan to do so by leveraging partitioning by company & date and reattach to the archive system, along with reference data
Challenges:
· The application is very big and using somewhat disjointed Data Access, so making wholesale changes is difficult
· The application makes heavy use of triggers and pseudo-materialized views (populated via triggers)
· Details on using partitions, especially on children table that currently do not have partition keys on them
· Our plan is to identify most frequently run/slowest queries and optimize to align to partitions
· Synchronizing Reference Data – How do we synchronize non-date based tables?
What we're looking for:
· Someone experienced with archiving patterns
· Partitioning and optimizing for its use (specifically dropping/add partitions and the impact.)
· Transportable tablespaces
· Reference Partitions and alternatives
· Identifying top sql to optimize to better align with partitioning
· Reference Data Synchronization – Batch or real-time. If real-time, assume Golden Gate..
Thanks,
Dhruv Soni
Contact : 980-272-1261
E-mail: dhruv@riderconsultinginc.com
Web: www.riderconsultinginc.com
0 comments:
Post a Comment