Gravity Servers: Data Centers and Locations
All of Gravity Servers’ data center facilities are located at Tier 3 and Tier 4 facilities with minimum cage level leases at every location and 24/7 staff on site. We have locations and network points-of-presence (POPs) in the United States and in Europe, and we are constantly looking for strategic locations to expand our footprint. All of Gravity Servers’ data center infrastructure runs on Cisco systems at all layers of the data center infrastructure with client distribution rack level switches and firewalls, top of rack, and end of row aggregation switches provided by Juniper networks.
Gravity Servers has space in four different US data centers as well as a data center in Zurich, Switzerland. All data server space is cage space, with staff on site, and all equipment owned directly by Gravity Servers. All data centers are fully SSAE16/SOC2 compliant with DWDM fiber networks. Power is provided via utility power, as well as backup UPS protected power and generator power if needed for long outages. Every facility has redundant cooling and all racks are laid out in a hot/cold aisle design with raised floors for maximum cooling efficiency. We can provide customer environments that are fully PCI DSS and HIPPA compliant, and have a dedicated Security Operations team on staff to assist with managed Firewalls, Intrusion Detection, and more.
The Gravity Servers Network
The Gravity Servers network is fully redundant, beginning at the gateway border routers down to the rack level switches. Every server rack, for instance, has a dual redundant switch with A+B fiber drops to assure that even in the microscopic chance of a dual switch failure and fiber failure, we would be able to roll to our redundant switch and fiber seamlessly. All firewalls, whether pulling overall network duty, or at the customer level, are always provisioned as a redundant pair (and done so at the customer level at no additional charge) to rule out single points of failure.
To reduce the risks of DDoS attacks and internal congestion, redundant 10-Gigabit Ethernet circuits are used across the entire network. Gravity Servers has been IPv6 compliant with full IPv6 connectivity in all data centers for two years as of 2014. The network is a fully switched network: each individual client is provisioned on its own VLAN, to prevent the packet sniffing of connections and traffic and to assure that every customer has their own private network.
Gravity Servers’ network is connected at various global points via 10-Gigabit Ethernet Wavelength handoff(s) on redundant/dual-path DWDM fiber-optic transport connections to Zayo® Bandwidth, Highwinds®, AboveNet®, Level3® Communications, and PCCW®. Transit providers include Level3® Communications, NLayer®/GTT, Zayo® Bandwidth, Highwinds®, AboveNet®, Cogent® Communications, and PCCW®. BurstNET® is currently accessible at public peering exchanges, including; NYIIX, Any2, LINX, NL-IX, & Equinix® Exchange.
Network Monitoring at Gravity Servers
Gravity Servers’ central Network Operations Center (NOC) is located in Delaware, and provides 24/7/365 monitoring of the entire network via Delaware as well as a world-wide monitoring cluster that has 15 locations. More locations allows us to determine if outages are localized or worldwide, and makes our NOC team capable of responding within minutes to server failures, attacks, or Internet backbone problems. As a result, the adverse effects of distributed denial of service (DDoS) and other attacks are minimized.
The Network Operations Center uses industry-standard SNMP (Simple Network Management Protocol) and provides round-the-clock monitoring of all hardware, including routers, switches, UPS systems, and servers. The Network Operations Center also monitors power, environmental factors (such as temperature and humidity), generator status, and network connectivity. All critical services/ports can be monitored, including FTP, HTTP, SMTP, HTTPS, SSH, and POP3. We use a mix of commercial and open-source monitoring products that include Nagios, Zabbix, MRTG, GroundWorks, custom scripts, and more.
Any failure or alert results in audible alarms and LAN messages, as well as email, pager, and SMS alerts. A fully automated escalation system is in place that notifies senior administrators and even management for alarms or outages that go over 5 minutes without being acknowledged, 10 minutes without an ETA to fix and diagnosis entered in the system, and 15 minutes without the problem being solved.
In addition, we regularly work with customers to deploy their own monitoring solutions with Open Source monitoring services such as Zabbix, ZenOSS, Nagios and others. As monitoring experts, we can work with you to achieve your own monitoring network that fits your needs at an extremely competitive cost. Many times hundreds of times under what other providers charge. We keep virtual machine images ready to deploy for customers with every major monitoring software preinstalled and ready to configure, thus saving you money, and us time. Contact our Sales department or your Account Manager for additional information on this service.
Gravity Servers DNS Cluster
Gravity Servers maintains its own five node redundant DNS cluster in the US with nodes in Seattle, Los Angeles, Dallas, New York and Jacksonville, Florida. Each DNS node is on a redundant, highly available setup at each location and does not exist within the Gravity Servers network. This means that in the event an entire data center goes dark, this would not affect DNS resolution at all. For the cluster to completely fail, a total of 10 servers at five different data centers, with two networks per data center would need to fail.
Having five locations also benefits DNS resolution time, making your sites and servers resolve quicker. Customers have total access to DNS records through our customer portal, including reverse DNS assignments. Our commitment to automation in this area pays off, allowing you, the customer, to completely manage your own DNS.
Updated on 2015-09-17T14:20:30-04:00, by .