always an important priority for
the VimpelCom team. This became
particularly important when
VimpelCom began to spread into
other regions outside of the
Russian Federation. “Our regional
expansion was contingent on our
ability to integrate acquired IT
systems quickly into a standard
architecture with established
policies and procedures,” Ilin
remembers. “In addition, availabil-
ity, security, and disaster recovery
requirements are baked into our IT
infrastructure. This is not always
the case with the IT systems of the
companies we acquire.”
But standardization goes much
further than delivering on promised
service level agreements (SLAs),
according to Ilin. “It also reduces cost
and complexity, and it provides us
with a flexible, scalable foundation
on which we can evolve our service
offerings to grow revenue and
market share in maturing markets,”
he explains.
Standardization key to
integrating acquisitions
In 2008, VimpelCom made two
strategic acquisitions—EDN Sovintel
and Corbina Telecom—that added
fiber to the premesis (FTTP), including IP-based television, network
services, to its product portfolio.
However, the IT infrastructures for
both of these companies were not
In 2008, former VimpelCom President
and CEO Alexander V. Izosimov published
an article in Harvard Business Review
in which he delineated five rules for
hypergrowth management: 2
1. Focus first on sales
2. Innovate with caution
3. Standardize structures
and processes
4. Delegate decisions to field
managers
5. Reward action and initiative
Managing Hypergrowth
as robust as those of VimpelCom
and moreover were complex and
based on different standards. “The
technology requirements and IT
infrastructures for Sovintel and
Corbina were completely different
than those of VimpelCom,” Ilin
observes. “For example, disaster
recovery and high availability were
real challenges for both, and neither
had many IT standards in place. We
also sought to use standardization
to reduce costs and complexity.”
However, due to other organizational exigencies and technology
challenges, the initial integration
efforts were put on hold.
“Over 80 percent of the data
now resides in our data center
environments and most of the
applications have been moved.”
IT infrastructure in bundles
The geographical dispersion of
VimpelCom’s operations requires
the company to maintain regional
computing assets—in terms of both
the telecommunications network
and the data center infrastructure.
The latter consists of a primary
location and secondary site located
about 10 kilometers apart. The
secondary location is about one-third the size of the primary site
“Availability, security, and disaster recovery requirements are baked
into our IT infrastructure.” – Dmitry Ilin, Director of IT Infrastructure, VimpelCom
When Ilin returned at the begin-
ning of 2010, one of his top priorities
was the integration of the Sovintel
and Corbina acquisitions. In particu-
lar, migration of storage from the So-
vintel and Corbina IT infrastructures
was one of the major undertakings
assigned to him. “It wasn’t simply a
matter of moving the storage assets
into our data center environments,”
Ilin states. “The companies both had
older equipment that was time-
consuming to manage and required
substantial maintenance costs.”
As a result, Ilin’s team retired
those systems and migrated the
storage to existing storage space
in their primary and secondary
data center environments. Veritas
Storage Foundation, which the
VimpelCom team has used since
the late 1990s, is being employed
to migrate storage from the legacy
Sovintel and Corbina IT storage
infrastructures.
“We’re nearly done with the
integration of IT systems from
Sovintel and Corbina,” Ilin reports.
and does not simply function as
a hot standby in the event of a
failure. Rather, the VimpelCom IT
team maintains an active produc-
tion and test environment in the
critical production systems. “This
allows us to drive hardware utiliza-
tion rates,” Ilin explains. “Hence,
we don’t have machines standing
idle and consuming valuable power
and cooling costs.”
At present, VimpelCom’s data cen-
ter infrastructure consists of about
1,000 Sun Solaris and 2,000 Microsoft
Windows and Red Hat Enterprise
Linux servers spread between the
primary and secondary locations
and another approximately 1,000
Microsoft Windows servers scattered
across the company’s seven regional
offices and 140 satellite offices.
Recognizing the importance of
meeting their SLAs, while reducing OPEX and CAPEX costs, the
IT team created two standard IT
infrastructure bundles. A high-performance bundle is comprised
of Sun Solaris servers and Hitachi
