Browse Prior Art Database

Database synchronization in embedded system

IP.com Disclosure Number: IPCOM000225600D
Publication Date: 2013-Feb-20
Document File: 3 page(s) / 22K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method for database synchronization in embedded system is disclosed.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 47% of the total text.

Page 01 of 3

Database synchronization in embedded system

Disclosed is a method for database synchronization in embedded system.

Some servers are equipped with an embedded computer known as the Service Processor (SP). Some of the higher end servers have redundant service processors to increase the system's RAS characteristics. The service processors can fail over to keep the server running in the event of failure of one SP. This means both SP including the back up or sibling SP data must be in sync at any given time to allow fail over. Thus, data synchronization is key to fail over to either redundant SP. Some of the critical data required for successful fail over are stored in SP databases. Since the service processor is an embedded system, it does not have the resource to handle a full fledged database servers that provide mirroring capability. Nevertheless, there is a need to keep the databases in sync.

The disclosed method provides a way to synchronize database data in an embedded system without hindering other database accesses during system run-time. When database data is changed on one of the SPs, the following two methods are disclosed:


1) Complete data synchronization


2) Immediate data synchronization.


1) Complete data synchronization: Some data does not need to be synchronized to the other SP until the SP boots the server to a certain point. For those type of data, complete synchronization is used. Complete synchronization means synchronizing full database table from one SP to another. This is done by reading table(s) data that needs to be synchronized from source SP and then merging/creating/updating it to the destination SP.


2) Immediate data synchronization: Some data is critical to the system and needs to be synchronized immediately. This is done by using update callback feature provided by SQLite. This callback is invoked by SQLite for any data change (insert/update/delete). Whenever this callback is invoked, same data is synchronized to the other SP. For the transaction level synchronization, start and end/rollback of the transaction is also synchronized along with the transaction data. Transactional level synchronization is done by providing special interfaces to start/end/rollback transaction. Any time this interfaces are invoked, same action is taken on the other SP database. This way if transaction is rollback on primary SP, it will also be rollback on other SP, thus keeping both SP data in sync.

Following steps are used for synchronizing database data from primary SP to redundant SP.

Pre-synchronization setup:


1) Whenever new database connection is opened, update callback is registered with SQLite.


2) Provide special interfaces for the start/end/rollback transaction.

1


Page 02 of 3

Complete Synchronization:

Complete synchronization involves synchronizing one or more tables from primary SP to redundant SP. Primary SP data is considered master copy of the data and thus full table(s) data from primary SP need to be synchroniz...