With the subscriber base of 146 million and still adding,
mention the rate at which the data is growing at Reliance
Communications (RCOM)?
RCOM has been growing consistently and hence the data management
practices are in place, benchmarked against the international
standards. We are constantly adding new customers every minute and
look forward to grow faster than ever. At the same time we offer
the VAS and 3G offerings.
Our data is growing enormously due to subscriber growth and
consumption of our services. We are applying innovative methods,
system and process changes to analyze, manage and make best use of
resulting data. This data has shown a growth of 50 to 55
percent.
How is the company managing this huge amount of
data?
RCOM uses a data management system, which monitors the data growth,
as desired. The data is segregated under three buckets:
- Customer centric – Required for customer services
- Business data – Required for analytics, trend analysis
and business forecast
- Legal data – Managed for regulatory requirements
We have configured our data storage system based upon the above
three criteria. This helps us analyze and manage the data
performance requirements, TCO. Our transition data goes into high
speed enterprise class storage, which has a very fast response
time.
The analytical kind of data is segregated into low cost enterprise
storage capable of high analytical performance. Multiple parallel
processing capable storage/application, with compression enabled is
used for legal data. This does not warrant high-end storage. We use
compression features of RDBMS, storage, applications to control the
data growth.
Is the company’s traditional storage system able to
handle this big data?
Yes, at present, thanks to our storage classification and
requirements we are able to make the ends meet. We use Tiered
Storage, which are a combination of fibre channels, SAS and SATA of
varied speed for different disk capacities. We also account the
storage fabric management cache and front-end adopters dedicated to
different kind of databases.
Process Improvement is a way of life at RCOM, we have undergone
database level structured revision and have removed obsolete
columns, which were the part of initial design basis. This has
helped lot, in storage optimization and reduction in future
requirements. As a part of our process improvement plan, we have
re-organised data layer to remove fragmentation, unwanted dumps and
unwanted data.
Is RCOM using any Big Data solution? If not, do you
have any plans to adopt it? Give reasons.
Due to sheer size of our databases, RCOM has been actively
involved in Big Data solutions. We have adopted GreenPlum (EMC) to
store CDRS and unstructured data and perform analytics on it. We
continue to test and evaluate other products like Vertica, IBRIX,
Hadoop etc. We have implemented storage virtualization. We have
also implemented unified storage (SAN, NAS) in a single box.
At present, what analytic solutions are being used by
the company. Does that suffice the need? If not, what are the
solutions the company is planning to implement?
RCOM uses SAS analytic, which is
software-based solution. We are also using TIBCO ODS and SAP BI
that help in analytic planning. For reporting purposes, we rely on
Business Objects. We also use Green Plum for its excellent analytic
capabilities.
As an innovative re-architectural process, we
are looking to re-architect the data management system by
classifying online and offline data. This will give us the maximum
benefit in data management and the way data is looked at.
About Author
Vinita Gupta is Principal Correspondent at InformationWeek India. Vinita has over six years experience in IT reporting and has interviewed more than 500 business executives. She has a PG Diploma in Business Management from NMIMS and Post Graduate Degree in Communication and Journalism from Mumbai University
Add description here
More articles by Vinita Gupta