|
Clients & Employers:
|
HP Labs
Southern California
Earthquake Center (USC)
Univ of Washington
Incorporated Research
Institutions for Seismology
Univ of Utah:
School of Computing
Computer Science
Campus Networking
Computer Center
RAHD Oncology Products
Univ of Utah Med Ctr:
Radiation Oncology
Diagnostic Imaging Lab
New York Univ Med Ctr
Radiology
Indiana Univ of PA:
Physics
Psychology
|
|
ATLAS
A High Availability Server
Designed Especially for Small Business
|
Most businesses rely on one or more servers to provide essential
services such as document management, database access, or email. When
a server is down, workflow processes fail. Productivity drops, and
profits are lost.
Large companies typically utilize High Availability Servers (HAS) to
minimize this problem. These, however, come with high costs that are
prohibitive for most small businesses. These high costs come in one of two forms:
|
|
1. The upfront cost for special hardware (nonstop computers).
2. The ongoing cost for necessary IT support (of failover clusters).
|
There has been no cost effective HAS for small business
until now.
ATLAS is designed specifically to fill this need.
|
Comparison of ATLAS and Existing Products
in terms of Failure Modes, Cost, and Function
Comparison of Failure Modes for All Options
| Failure Mode |
Hot Swap Server |
Non-stop Computer |
HA Cluster |
ATLAS |
| Power Supply Fails |
UP | UP | UP | UP |
| Fan fails |
UP | UP | UP | UP |
| Disk fails |
UP | UP | UP | UP |
| CPU fails |
DOWN | UP | UP | UP |
| Memory fails |
DOWN | UP | UP | UP |
| Motherboard fails |
DOWN | UP | UP | UP |
| I/O Bus fails |
DOWN | UP | UP | UP |
| Disk controller fails |
DOWN | UP | UP | UP |
| Net Interface fails |
DOWN [1] | UP | UP | UP |
| Other hardware fails |
DOWN | UP | UP | UP |
| Planned maintenance |
DOWN | DOWN | UP | UP |
Notes referenced above:
- Can be UP if possible to configure with multiple NICs and failover.
Comparison of Cost and Function of High Availability Options
| Cost (Function) |
Non-stop Computer |
HA Cluster |
ATLAS |
| Cost (system) |
High | Low | Low |
| Cost (IT resources) |
Low | High | Low |
| Type of failover |
Auto | Auto | Manual [3] |
| Danger of split brain? |
No | Yes [2] | No |
| Confidence at failover? |
Yes | No [1] | Yes |
| Verification of services? |
No [1] | No [1] | Yes |
| Synchronization? |
Yes | No [1] | Yes |
| Hardware failures |
UP | UP | UP |
| Planned Maintenance |
DOWN | UP | UP |
| Load sharing possible? |
No | Yes | No [3] |
Notes referenced above:
- Possible only if the end user provides for this.
- Potential loss of data.
- These increase confidence at failover.
|
To learn more, see selections
under High Availability Server
at left.
|
|
|
