Physically Secure Environment
Prominic.NET tightly controls physical access to all areas of the Prominic.NET Data Centers. Only highly trained Prominic.NET personnel are allowed access to the data center floor and interior server racks. Access to the data center building is controlled by combination key card/code panel, and personnel must pass through a second locked area in order to access the data center floor. Visitors are not allowed in the data center without a Prominic.NET escort and we discourage site visits for sake of security. The entire data center is alarmed against intrusion, and both Prominic.NET and local police are notified in the event an alarm is tripped—at any time and for any reason.
The Prominic.NET Data Center "core" runs on Gigabit Ethernet, connected to the Internet through multiple redundant high-speed Internet lines using high-availability Cisco routers. Our transit providers include Paetec, Sprint, Comcast, and Atlantic Metro for Internet connectivity. We continuously explore relationships with additional Internet backbone carriers for extra redundancy, including our participation in the CHIX Chicago Internet Exchange.
We always guarantee plenty of excess Internet capacity so you never have to worry about your bandwidth. Unlike many providers which boast impressive sounding numbers such as "multiple OC-12 connections" without revealing actual remaining line capacity, we guarantee a maximum average utilization of 70%. This means our metric for Internet line upgrades is to have 30% excess capacity during peak load times, regardless of how many high-speed lines we already have. High-speed lines are not useful to customers when they are full—and unlike other providers we address this problem as a core business policy.
Power, Alarm & Environment
Rack mounted servers and networking equipment are housed in a physically secure environment (personnel-restricted areas) with multiple alarm sensors alerting us for intrusion, smoke, fire, and excess temperature. Equipment areas have plenty of Uninterruptible Power Supply (UPS) capacity for minor power interruptions and our natural gas generators will run indefinitely in the event of a failure in the power grid. Furthermore, multiple redundant environmental control units provide special computer-room grade cooling. This ensures that all of your mission-critical electronic components are always well within their specified operating conditions.
Firewall Protection—Free, Because It's That Important
Most data center providers charge extra to have your server protected by a firewall. We believe firewall protection is too critical to be optional under any circumstances, so we include this feature with all of our hosting packages. We rely on ICSA-certified firewall technology provided by Global Technology Associates (GTA), makers of a high availability firewall system called GNAT-Box.
The GNAT-Box, unlike many other firewall systems, is a completely self-contained miniature operating system and full Network Address Translation firewall with stateful packet inspection. The advantage of this architecture over traditional Windows- or UNIX-based firewalls is that GNAT-box is not susceptible to the security flaws of an operating system outside its control. The GNAT-Box's small footprint, secure TCP stack, and miniature UNIX-like kernel ensure total understanding of both firewall and operating system functions by the firewall programmers. This means a more reliable, more secure and faster firewall than any big-footprint firewall vendor can provide.
All servers are located behind a firewall at no additional charge, and for each server, we only open the necessary ports for the server to function, such as 1352 for IBM Lotus Notes client access, 80 for web access, and other ports optionally for POP3, SMTP, SSL, etc. Dedicated Firewalls with VPN and high-availability capabilities are available to dedicated server customers for an additional charge.
24 x 7 Server Monitoring
24 x 7 x 365 uptime monitoring is provided by our custom network monitoring system. All critical systems are monitored constantly, and alerts are sent out by e-mail and pager immediately after problem detection.
Our monitoring system detects and alerts us to errors immediately as they happen. Simultaneously, our system recovery architecture kicks in to immediately self-heal the network and application software. This system solves 99% of all routine hosting problems before a human operator even has the chance to intervene. This means you get the highest uptime possible!
Servers are monitored 24/7 on a number of different levels. We monitor the status of the Internet lines, the health of our internal network in general, and we also monitor the individual servers. On each server, we monitor the Domino server task itself and also the HTTP task running on the Domino server, since it can become unresponsive even when the rest of server is still running smoothly. Servers are polled every minute, and if a problem arises for longer than 2 cycles (2 minutes), an administrator is e-mailed as well as electronically paged. If the problem is not resolved within 25 minutes, the original administrator is paged along with a second group of administrators.
For our dedicated customers, we are able to provide a special active monitoring technology that we have developed to make our own servers self-healing. In the event of a Domino server crash or a server becoming unresponsive, our monitoring system can kill the server process, flush its memory, and restart the Domino server without affecting the machine as a whole. This is important for partitions because we can reset one partition on a machine without negatively impacting the other partitions on the machine. Equally important, dedicated machines that have non-Domino services can continue to run uninterrupted while the Domino server is restarted. The end result is that you experience maximum uptime without the added cost of human intervention.