Network information server

Details Network information server Network information server

1998-05-01

2001-07-10

Trammell, James P. (Department: 2131)

Error detection/correction and fault detection/recovery

Data processing system error or fault handling

Reliability and availability

C714S046000

Reexamination Certificate

active

06260155

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a high speed, microprocessor based fault tolerant mass storage information server system, and more particularly concerns an integrated, fault tolerant, four-way mirrored Internet server system, including hot swappable components connected to a midplane connector, combining communication, routing and switching functions with scalable storage, and an integrated system monitoring diagnostics computer subsystem.
2. Description of Related Art
Conventional information server systems, including Internet server systems, have been able to provide only limited scalability to accommodate increased storage requirements, and have typically required a power-down of the information server system to facilitate installation of additional disk storage devices or communications components. The shut down and installation of additional or replacement components to an information server system commonly requires a significant amount of system administrator personnel time and resources to facilitate system maintenance, and for such tasks as expansion of storage space, repair, and routine maintenance, such as optimization and system monitoring.
Conventional information server systems also have limited fault tolerance, and typically employ a disk array server which incorporates extra disk space substantially in excess of that needed by the information server system. The extra disk space is incorporated into such systems by the addition of disk storage devices. However, conventional backplane technology typically has been limited to allowing up to eight disk drives in an individual server. Multiple server units could be linked to increase the capacity of an information server system, and this type of multiple server configuration also allowed only a limited amount of fault tolerance. Typically, if one server failed, information directed to the server would be directed to the next server in line for the registration, and so on from one server to the next. However, each time the information was rerouted, timing out was required, resulting in compounding of delays as information was transferred from one server to another. Prior art systems also balanced the load among clustered servers by utilizing a load balancing program, which, however, required an operator to manually run a diagnostics program on each server and manually switch servers.
Depending upon the nature of a component failure, corrective action with prior art systems can include removal and replacement of a completely or partially failed device. Such system maintenance typically requires that the entire server system be taken off-line, and powered down before the physical replacement operation can be performed. Some information server systems permit “hot swapping,” the removal and replacement of a component in a system while the power is on and the system is operating. Such systems still typically require considerable hands-on intervention of the system administrator personnel to manually manipulate the hardware interfaces and operating system software for purposes of physically and logically reintegrating the newly replaced disk device or other component into the server system.
The process of logical reintegration of even a “hot swappable” component into a server system typically requires a significant portion of the CPU and memory resources of the server system. For installation of a disk storage device, such resources are needed to reallocate the available free disk space into the logical disk configuration of the server, and to redistribute the multiple copies of the information stored on the server across the new and remaining physical and logical disk drives. Such demands on the resources of the server system can temporarily severely decrease the performance of the information server system. There is therefore clearly a need for a fault tolerant information server system that allows for the rapid removal of failed or failing components, and the integration of either replacement components or additional components into the server system, while the power is on and the system is operating, without a significant drain upon the resources of the server system.
With reference to
FIG. 1
, illustrating a prior art server hardware configuration, a conventional information server system includes a variety of components from different manufacturers and utilizing different protocols, frequently causing incompatibility problems in performing diagnostics of the entire server system.
Such a server system typically is connected to receive user calls, forwarded to the server system by a central office (CO) through a normal telephone line (POTS) carrier line to a modem, such as a modem available from U.S. Robotics, for example, interfaced with a terminal server such as those provided by Livingston, for example, and a hub for connecting several computers or networks together, such as those provided by Kingston, for example. The hub is connected to a communications system, such as that provided by CISCO, and a router such as that provided by ADTRAN, which is in turn interfaced to the network (WWW).
Such prior art network interface systems have serious limitations in their adaptability and flexibility, and must be directly managed by skilled personnel who are knowledgeable about how to effectively maintain availability of the network interface in the face of the complex relationships of the hardware used. Often, the only way to achieve reasonable reliability is to provide large amounts of redundancy as well. All of the above contributing to a great deal of expense in the requisition and maintenance of a network interface system. The present invention solves these and many other problems.
SUMMARY OF THE INVENTION
Briefly, and in general terms, the present invention provides for an information server system that provides a wide array of reconfigurable options as well as scalable expansion capability, due to the implementation of a unique mid-plane connector board for all components of the information server system. The information server system can thus be readily assembled and integrated with multiply mirrored components to significantly improve fault tolerance of the server system. In addition, the information server system of the invention provides an integrated diagnostics computer subsystem and fully hot swappable components that eliminate problems of degradation of server system performance by replacement or addition of components to the server system. In addition, the integrated diagnostics computer subsystem eliminates the incompatability problems previously associated with performing diagnostics on the server system. The rate of end user communications reception can also advantageously be increased by a method for installing a hash table database in the end user computer system to bypass a similar but slower function in the end user modem, to place the bulk of the transmission reception load on the main CPU, and synchronizing the modem with the transmission speed of the server unit.
The invention accordingly provides for an information server system having a scalable, modular, fault tolerant, hot swappable architecture of a plurality of components for interfacing with a computer network such as the Internet, comprising a central processing unit, a communications interface subsystem connected to the central processing unit, a mass storage subsystem connected to the central processing unit, and a diagnostics computer subsystem connected to the central processing unit for monitoring when one of the plurality of components of the information server system is removed or added. In a presently preferred embodiment, the information server system further comprises a mid-plane connector board having a plurality of sockets for connecting the components on each side of the mid-plane connector board, and the components of the information server system are currently preferably connected to the mid-plane connector board of the information server system by interface cards. In a preferred

Affiliated with

Also associated with

Rating

Network information server does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Network information server, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Network information server will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2469927