Web Hosting Data Center
Today's competitive e-business marketplace requires that your business
uses the most efficient and reliable hosting services. As client of Dhaka-Hosting
on the Internet today, you will reap all the benefits of our
state-of-the-art data center.
The physical environment necessary to keep the servers up and running 24
hours a day, 7 days a week is provided by Hostcentric's Internet Data
Centers. Dhaka-Hosting uses a data center located on the West Coast of the
USA. The building is a class A, state of the art facility data center
which is earthquake proof.
High Bandwidth Connectivity:
Our Data Center is owned by Hostcentric, located at Fremont, CA. The
network capacity exceeds 2.5 Gigs of Internet connectivity via DS3, OC3,
OC12, and Gigabit Ethernet connections to a wide variety of peer points
and several transit providers. The circuits are terminated in our data
center on carrier class, Cisco Systems 12000 routers.
Power quality in the datacenter is ensured by multiple independent Liebert
UPS systems. Should the utility power fail, the mission critical
electrical loads at the datacenter are provided by the Liebert UPS
systems, which are configured with automatic static bypass and manually
operated full-maintenance bypass circuits. Each UPS module has its own DC
battery bank with sufficient capacity to sustain its critical bus for
periods exceeding 20 minutes without additional power supply from utility
or generator sources.
Standby generator: The electrical utility service (PG&E) is backed up
via a stand-by diesel power generator activated by an Onan automatic
transfer switch. The generator is a 750 KW Cummins-Onan, with sufficient
on-site fuel to run continuously for over 24 hours.
The Dhaka-Hosting network is a fully switched network. Traffic destined
for a website hosted on a server is sent only through switching equipment
necessary to deliver the information to that server. This minimizes the
risk of someone sniffing or capturing traffic being sent across the
Our data center, was built from the ground up and is equipped with various
high capacity connections to multiple carriers to ensure uptime and speed
consistency and redundancy.
The data center's HVAC (Heating Ventilation Air Conditioning) system is
N+1 redundant. With full particle filtering and humidity control, the
environment is maintained at a cool 68° F to ensure a comfortable
environment for our servers.
As a Tier 1 ISP, Hostcentric has established extensive public and private
peering relationships. Peer points currently include the Pacific Bell NAP
and Palo Alto Internet Exchange. BGP4 is used for optimal route selection
and automatic fail over.
Network bandwidth is carefully monitored to ensure that customer
utilization does not exceed online capacity during peak Internet traffic
Our datacenter "Cisco Powered Network" relies on redundant Cisco
12000 series routers and 6500 series switches at its core, and can be made
fully redundant all the way down to the customer server.
Dual core routers at the top layer connect to a layer-three, switched
backbone. This switching fabric connects to a layer two distribution
switch infrastructure that in turn connects to a bandwidth managing
"Officer". The bandwidth manager controls traffic from multiple
layer two switches mounted on the customer's rack that are connected to
individual servers. For additional redundancy, customers may be dual homed
by purchasing an additional Ethernet port and diverse cabling that
terminates on a second switch.
Hostcentric Officers provide traffic graphing and sophisticated
bandwidth management. To monitor their packet traffic customers are
provided with one bandwidth graph per Internet uplink connection.
Hostcentric network operation centers are staffed 24 hours per day, 7 days
each week by skilled technicians. Industry-leading service level
agreements highlight company guarantees like 100% network uptime.
The servers we are using are Dual P4/Xeon at least 2.6 GHz with 1.5 or 2
GB memory, SCSI hard drives in RAID and backups. Each server is using a
gigabit network card.
We keep spare
servers on-line of all CPU configurations. If a server were to experience a
hardware failure, we would turn a key, grab the handle on the drive, pull it
out, and insert it into an identical standby CPU. We would then reboot the
second machine and the server would be up and running again in a matter of