In a system for appliance back-up, a primary appliance is coupled to a network, whereby the primary appliance receives requests or commands and sends a status message over the network to a standby appliance, which indicates that the primary appliance is operational. If the standby appliance does not receive the status message or the status message is invalid, the standby appliance writes a shutdown message to a storage device. The primary appliance then reads the shutdown message stored in the storage device and disables itself from processing requests or commands. When the primary appliance completes these tasks, it disables communication connections and writes a shutdown completion message to the storage device. The standby appliance reads the shutdown completion message from the storage device and initiates a start-up procedure. This procedure causes the address of the standby appliance to be identical to the primary appliance address, and the standby appliance processes the requests or commands in place of the primary appliance.

System and method for fibrechannel fail-over through port spoofing

When updating a running program in a system that uses a one-to-one backup program, fault-tolerance is lost while the backup program is itself being updated. To overcome this temporary loss of an available backup program, the number of backup copies of a software process is temporarily and dynamically increased during the software upgrade. The extra backup software processes may run on an unused processing unit or may run as an extra software process on a processing unit which is already performing a task. The technique may be applied to communication line cards.

Method for upgrading running software processes without compromising fault-tolerance

A method of processing a transaction includes processing a transaction workload in a primary process pair on a first node in a cluster of nodes, the processing using at least one stable storage volume for storing a database and another stable storage volume for storing a log, the at least one stable storage volume and the log storage volume forming a log storage group. The method further includes performing checkpointing operations via the network from the primary process pair to a backup process pair while processing the transaction workload, the backup process pair operating on a second node in the cluster of nodes. The method further includes detecting a failure making the first node inoperable or inaccessible, and after detecting the failure, engaging the backup process pair to take over the transaction processing workload of the primary process pair, the backup process pair being configured to operate with the log storage group used by the primary process pair on the failed node.

Transaction processing apparatus and method

A distributed software redundancy design is disclosed to minimize network outages and other problems associated with component/process failures by spreading software backup (in the so-called “hot state”) across multiple elements. The distributed redundancy architecture of the present invention also permits the location of the hardware backup element to float, that is, if a primary element fails, the functions can be transferred over to the backup element. When the failed primary element is replaced, the replacement hardware can serve as the hardware backup. If one or more of the primary processes on a particular element experiences a software fault, the processor on the line card may terminate and restart the failing process or processes. Once the process or processes are restarted, a copy of the last known dynamic state (i.e., the backup state) can be retrieved a from corresponding backup processes executing on a second line card and initiate an audit process to synchronize retrieved state with the dynamic state of associated other processes.

Distributed process redundancy

This disclosure describes a technique including operating a redundant disk drive device, wherein the redundant disk drive device includes a primary disk drive and at least one mirror disk drive. At least one of the disk drives is reconfigured within the redundant disk drive system in such a manner that during the reconfiguring, the disk drive is maintained on-line.

In-service raid mirror reconfiguring

In a multiple processing system or cluster, a pair of processes, assuming the role of a primary process and a backup process to the primary process, are replaced on-line by stopping the backup process; creating the replacement backup process; checking to ensure compatibility between the primary and replacement backup processes so that communication between them is possible and information sent by the primary process to the replacement will be correctly received and handled; providing the replacement backup process with that state of the primary process needed in order to take over the function and operation of the primary process; switching roles so that the replacement backup process now takes over the function and operation of the primary, and the primary becomes the backup; and repeating the steps of creating, checking, providing and switching, to conclude with a newly-installed replacement primary process and a replacement backup process, completing the on-line replacement of the process pair.

On-line replacement of process pairs in a clustered processor architecture

A fault tolerant cluster of computer systems includes a “process quad” comprising four duplicate processes—a primary process and a backup process on a primary system, and a primary process and a backup process on a backup system. The state of the backup process on the primary system is maintained by receiving checkpoint information from the primary process on the primary system, and the states of the primary and backup processes on the backup system are maintained by receiving checkpoint information either directly or indirectly from the primary process on the primary system.

Using process quads to enable continuous services in a cluster environment

A memory node for fault tolerant computing includes a log pipeline coupled to a compute node. The memory node includes an active log having multiple log entries received from the log pipeline of changes to a set of transactional memory pages in the compute node, and a set of active pages. The memory node processes the active log to keep the set of active pages in one of a fully recovered fresh state and a stale state reflecting within a predetermined delta of log entries of a corresponding transactional memory page where heavily accessed and critical pages are acted on before other pages.

Wendy B Bartlett

Papers

On-line replacement of process pairs in a clustered processor architecture

Using process quads to enable continuous services in a cluster environment

Fault tolerant computing