The Cbeyond networking team will work with you to determine what method of load balancing is best based on the needs of your individual business.

Cbeyond load balancing powered by F5 BIG-IP Local Traffic Manager offers a valuable advantage for two or more physical servers. The hardware load balancing allows Cbeyond to provide high-availability services based on server fail-over, traffic load, or simple round-robin approaches. This optimization between servers ensures that the web surfer receives the fastest possible performance with the shortest download time.

In addition to load balancing the benefits for customers include:

• Increased application capacity
• Improved scalability
• Increased reliability by distributing total traffic loads
• Redistributing the load of a failed web server
• Providing proactive server health checks to determine availability

Cbeyond delivers the most cost effective load balancing solution to meet the needs of all of our customers at both ends of the user traffic spectrum - all at an affordable price with proven performance.

Round Robin

This is the default load balancing method. Round Robin mode passes each new connection request to the next server in line, eventually distributing connections evenly across the array of machines being load balanced. Round Robin mode works well in most configurations, especially if the equipment that you are load balancing is roughly equal in processing speed and memory.

Least Connections (member) and Least Connections (node)

The Least Connections methods are relatively simple in that the LTM system passes a new connection to the node that has the least number of current connections. Least Connections methods work best in environments where the servers or other equipment you are load balancing have similar capabilities.

These are dynamic load balancing methods, distributing connections based on various aspects of real-time server performance analysis, such as the current number of connections per node or the fastest node response time. Load balancing calculations may be localized to each pool (member-based calculation) or they may apply to all pools of which a server is a member (node-based calculation).

Specifying priority-based member activation

You can load balance traffic across all members of a pool or across only members that are currently activated according to their priority number. In priority-based member activation, each member in a pool is assigned a priority number that places it in a priority group designated by that number.

With all pool members available (meaning they are enabled, marked up, and have not exceeded their connection limit), the LTM system distributes connections to all members in the highest priority group only, that is, the group designated by the highest priority number. The Priority Group Activation value determines the minimum number of members that must remain available for traffic to be confined to that group. If the number of available members in the highest priority group drops below the minimum number, the LTM system also distributes traffic to the next higher priority group, and so on.

Load Balance Pool Monitors

Monitors are how the load balancer knows which real server is available to accept connections. The default monitor is setup to monitor ports 80 and 443 to see if these ports are still active on the real servers. If they are no longer active then the load balancer will mark that pool member as down and stop forwarding connections to that server. Once the ports are up again the load balancer will start sending traffic to that server.

There are also custom monitors that can be setup to monitor specific URLs. This is beneficial if an application goes down, but port 80 is still responsive. The load balancer will try to hit port 80 first and it will be successful, but when it tries to open the URL and it is down the load balancer will then stop forwarding connections to that particular server.

Special order requests for load balancing methods are also available. For more information about load balancing methods and special order requests, please contact a Customer Solutions Advisor through the Cloud Server Order Request Form or contact us via telephone or online chat.

^ Return to Top

Clustering is a hardware-based solution that relies on high performance fiber optic connections to a fully-redundant SAN storage solution. 4Gb infrastructure with dual fiber HBAs actually increases performance beyond local disks. Failover happens within seconds for the typical database, however extremely large databases will increase the necessary time. Two servers in an active-passive configuration share the same storage, with the SQL instance failing to the passive server in the case of failure. Transactions seamlessly move to the active server so that no changes to the database are lost. Cbeyond includes the SQL Support package for clustered solutions.

This process copies live transactions to a secondary server with automatic failover in the event of primary server failure. There is a performance loss with mirroring, as every transaction effectively has to be written twice, but since it does not require the same level of additional hardware and infrastructure, it is a more affordable High Availability solution. Mirroring is configured at the database level, so the customer must notify the SQL team when new databases or users are added, so that they can be included. Also, .NET is required to fully utilize this solution as a back-end to a website.

Log Shipping is the process of backing up a production database transaction log, then copying the backup to a second server. The backup is then restored on the second server and placed in a recovering state, typically every 15 minutes. A DNS alias is created for the customer to connect to the production SQL Server, and in the event of a failure is pointed to the standby server, which is placed into an active state. In this scenario customers must contact Cbeyond if they add databases or users. A failover takes about 15 minutes to perform, and is not automatic. This is a disaster-only solution.

^ Return to Top 

Cbeyond utilizes the Very Early Smoke Detection Apparatus (VESDA) to sniff the air within the data center zone in which the unit is deployed. Using a set of tubes to pull air from nozzles within the space, the VESDA system uses aspiration to sample the air for early warnings of an impending fire.

The sensitivity of the VESDA systems allows multiple levels of alert to be sent to the data center staff based on the severity of the detected threat. This tiered system allows the staff to make a visual inspection in low-threat situations, make a manual deployment of the fire suppression in mid-level threats when deemed necessary, or trigger an automatic deployment of the suppression agent in a high-threat detection.

Our data center is partitioned into 2,500 SQ FT rooms in a zoned approach. This is to insure if a fire is started in one zone no other zones are effected and the fire can be contained to the single zone. Due to the added cost and complexity of zoning distributed cooling, power, and suppression systems, very few data centers are designed with this extra measure of safety.

Each zone has its own separate FE25 clean agent fire suppression system for continuous operation of non-effected equipment in the event of a fire. This fire suppression system is fully managed and maintained by one of the best fire suppression companies in the world, one of the first companies to obtain their ISO-9001 certification.

• Fully managed state-of-the-art fire suppression system
• Managing company is ISO-9001 certified
• Fully isolated 2,500 SQ FT fire suppression zones
• The latest fire suppression technology using FE-25

Equipment within the data center is provided with conditioned electrical power via two redundant 625Kva Liebert UPS systems. These UPS's provide both active power conditioning to control voltage sags and spikes, and emergency power in the event of a loss of utility power. These units are a critical component of the always-on power supply to your equipment as they provide power during the interval between loss of utility power and the firing and synchronization of the diesel generator.

Self-power is provided to the data center in the event of a utility outage through the use of a two megawatt diesel generator and Automatic Transfer Switch (ATS). The generator selected for emergency power needs has been sized with future growth in mind and easily provides ongoing emergency power through the use of a 4,500 gallon base fuel tank and a standby refueling contract. With sixteen cylinders, four turbochargers, and weighing in at 58,000 pounds, this is a very capable source of backup power. To ensure that the generator is always ready for immediate service, a preset test-firing sequence is executed once each calendar month.

Our data center uses a Liebert cooling system that utilizes four 30 ton air handlers in each 2,500 square foot data center zone, each with redundant compressor units and integrated humidity controls. These internal units are supplied by three 50 horsepower coolant pumps pushing supply and return fluids to three external cooling towers. This system provides excellent redundancy, scalability, and environmental management.

^ Return to Top 

What policies and procedures are in place for Physical Security?

• The data center floor utilizes the latest VESDA smoke detection and FE-25 clean agent fire suppression for uninterrupted operation of non-affected equipment in the event of a fire.
• All off-site data storage is maintained in a specialty built Fire King DS-Series data safe, utilizing electronic and mechanical locking mechanisms.
• Biometric and employee specific pass codes are required to pass through all security checkpoints.
• All internal and external entry points utilize reinforced steel doors and door frames, safety glass, and walls. These access points are video recorded and archived for one year.
• Facility is wired with glass break sensors and visual and audible alarms to alert staff in the event of an external glass break.

^ Return to Top

Contact a Customer Solutions Advisor through the Cloud Server Order Request Form and receive a price quote within 2 business days.