System and method for synchornisation of data and recovery of failures during synchronization between two systems

• US 9,262,282 B2
• Filed: 09/03/2013
• Issued: 02/16/2016
• Est. Priority Date: 09/04/2012
• Status: Active Grant

• Current Assignee: Opshub, Inc.
• Original Assignee: Opshub, Inc.
• Inventors: Jain, Sandeep, Tiwary, Prakash Chandra, Garg, Aparna
• Primary Examiner(s): PATEL, KAMINI B
• Application Number:
  US14/017,244
• Publication Number:
  US 20140068328A1
• Time in Patent Office: 896 Days
• Field of Search: 714/15, 707/610, 707/615, 707/616, 707/618, 707/641
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1443 Transmit or communication errors
• CPC Class Codes: G06F 11/1471 involving logging of persistent data for recovery
• CPC Class Codes: G06F 11/2079 Bidirectional techniques
• CPC Class Codes: G06F 11/2082 Data synchronisation
• CPC Class Codes: G06F 11/3676 for coverage analysis

Abstract

The embodiments herein provide a system and method for synchronization of data and recovery of failures during synchronization between two systems. The system for executing synchronization between the source and target systems comprises a connector framework, a process framework, a mapping manager, a recovery manager, a conflict manager and a failure manager. A method for bidirectional data synchronizing and integrating involves keeping information about integration user in integration database for each entity and update along with update time and audit id, only when the system does not keep information about the created and updated entity. A method of synchronizing only the entities qualifying a criterion, a method for failure recovery from a correct point during integration process, a method for handling recovery for changes coming from multiple end points and a method for recovery during server shutdown are also provided.

Claims

1. A method for synchronization of data and recovery of failures during a bidirectional synchronization between two systems, the method comprises the steps of:

polling a right data during a bidirectional integration between a source system and a target system;

synchronizing a new create from the source system to the target system;

creating an entry in the target system, wherein an entity is filtered out when the entity is created by a same integration user during a polling of the entity;

saving the created entry to the integration user;

generating a unique global id associated with each entity;

detecting conflicts during the bidirectional synchronization and when a data is moved in one way from the source system to the target system;

generating the global ID for each entity based entity type and a system internal id by the integration manager;

searching an associated global ID in the target system by the integration manager;

fetching and writing a target information, when a matching entity is found in the target system based on generated global ID and the searched associated global ID by the integration manager;

generating a new global id and storing the generated global ID in an integration storage, when a matching entity is not found in the target system by the integration manager; and

updating the target information in a table by the integration manager;

detecting conflicts during a bidirectional synchronization and when a data is moved in two ways from the source system to the target system and vice versa;

recovering from a system failure; and

recovering from an application failure.

2. The method of claim 1, wherein the method further comprises the step of detecting conflicts in the unidirectional synchronization, said method comprising the following steps:

creating an event (E) with a state (S1) in the source system,

synchronizing the event from the source system to the target system, and wherein the event (E) is replicated to the target system with the state (S1);

updating the event (E) in the target system and changing the state (S1) to state (S2) after updating;

updating the event (E) in the source system with a state (S3);

synchronizing the event (E) from the source system to the target system,

and wherein the event with (S3) is written to the target system; and

Detecting a conflict by identifying a change in the state from (S1) to (S3) for the event (E).

3. The method of claim 2, wherein the step of detecting conflicts in the unidirectional synchronization comprises the steps of:

creating an event (E) with state (S1) in the source system;

synchronizing the event (E) from the source system to the target system, and wherein the event (f) is replicated to the target system with state (S1);

updating the event (E) in the target system and changing the state from (S1) changed to (S2);

updating the event (E) in the source system with a state (S3);

synchronizing the event (E) from the target system to the source system with the event (E) with state (S2);

detecting a conflict by identifying a change in the state (S3) for the event (E) by an integration, and wherein the conflict occurs in an integration, when old state of the source entity is not same as the current state of the entity at target system, and wherein the conflict is detected by comparing the old state of the source entity with the current state of the target entity;

synchronizing the event (E) with state (S3) from the source system to the target system; and

detecting conflict by identifying change m state (S2) for event (E) by the integration.

4. The method of claim 1, wherein the step of recovering from the system failures comprises the steps of:

storing a time stamp of the source system and a step id that are last processed for each entity by the integration system for handling failures;

resuming the synchronization from an exact point of failure; and

allowing updating of further events whose time is greater than or equal to the time of source system and when a step id of the event is greater than 1.

5. The method of claim 1, wherein the step of recovering from the application failures comprises the steps of:

sending the event by source system pollers;

assigning a constant step number to each sub step in an update process;

getting a last updated time value from the target system for the entity that is updated;

updating the integration storage with the event and a respective last storage time;

checking fix changes done by the integration user after a last storage time for the event along with a time of changes, before initiating an update;

updating the time of changes in the integration storage, when the changes are found; and

creating/updating the entity in the integration storage, when the changes are not found.

6. The method as claimed in claim 1, wherein the method further comprises:

synchronizing preselected entities which qualifies a preset criteria;

assigning a flag to the preselected entities, when the preselected entities are updated after a time of first poll and the changes to the preselected entities are not synchronized to the target system in next time as the preselected quantities fail to satisfy the criteria due to updating; and

synchronizing the entities with flag-on condition.