Method and device for database & storage aware routers

Patent metadata

Field	Value
Publication number	US-2016321341-A1
Application number	US-201615148910-A
Country	US
Kind code	A1
Filing date	May 6, 2016
Priority date	Nov 8, 2013
Publication date	Nov 3, 2016
Grant date	—

Abstract

A method and system for optimizing replication in a distributed network is described. The instant invention allows for determining existing cluster topology of the network by one or more router(s) device(s) operating in the network, identifying and optimizing a data replication stream/service in use in network, by said router(s); determining a routing scheme based on the cluster topology by the router routing data packets though said network based on said routing scheme and applying predefined policy to a predefined set of router(s) corresponding to identified data replication stream by the router.

First claim

What is claimed is:

1 . A method for optimizing replication in a distributed network configured for hosting and communicating among plurality of data packets and a database server, the method comprising:
determining existing cluster topology of the network by a router operating in the network; identifying and optimizing a data replication stream in use in the network by the router; determining a routing scheme based on the cluster topology by the router; routing data packets through the network based on the routing scheme; and applying a predefined policy to at least one predefined router corresponding to a identified data replication stream by the router.

2 . The method as claimed in claim 1 , wherein the router comprises:
at least one edge router and an edge router operating closest to the database server is a database node edge router.

3 . The method as claimed in claim 2 , wherein the database node edge router includes configured data comprising:
an Internet Protocol address of the database node edge router; a database cluster unique identifier; and channel and associated policy configurations.

4 . The method as claimed in claim 2 , wherein a master edge router optimizes data replication stream by enabling caching on the at least one edge router and attaching a fingerprint to a payload associated with the cache.

5 . The method as claimed in claim 1 , wherein routing data packets comprises:
creating a path along with a label by a first router for routing a data packet to a third router using cached data of a second router; determining by the first router that a data packet to be transmitted to the third router is cached in the second router and using a cached finger print along with the path label; and receiving by the second router the fingerprint with corresponding label, processing the received packet data, converting the packet data with actual data and forwarding the actual data towards third router.

6 . The method as claimed in claim 1 , further comprising:
advertising/publishing the identified data replication stream by the at least one edge router in the network; receiving and storing the advertised/published information by a database-aware router; subscribing to a master database edge router by the database-aware router; adding the database-aware router in a cluster as a SDER (Subscription Database Edge Router) after subscription; configuring the corresponding SDER for each added database-aware router; and wherein the master database edge router (MDER) on receiving a subscription request is configured for:
checking if requesting database-aware router or SDER is already subscribed, and
enabling caching on MDER-SDER link after determining that requesting database-aware router or SDER is already subscribed.

7 . The method as claimed in claim 1 , further comprising:
sending REDO log information as a data packet to a subscriber node by its corresponding master DB node; intercepting the packet data by a MDER (master database edge router) for determining REDO log information; caching the packet data after determining that it contains REDO log information and that cache is not already enabled.

8 . The method as claimed in claim 7 , further comprising:
extracting destination address of a data packet by the master database edge router via a lookup edge router information; generating a finger print for the packet data for a cache enabled edge router, and storing the fingerprint; labeling the data packet and forwarding the fingerprint; and generating a label by the master database edge router for the packet data and forwarding only the finger print when it is determined that a fingerprint already exists.

9 . The method as claimed in claim 1 , wherein a master database edge router based on a determined topology of a DB cluster exhibits an optimal utilization of caching.

10 . A system for optimizing replication in a distributed network configured for hosting and communicating among plurality of data packets and a database server, the system comprising:
one or more routers operating in the network and configured to:
determine existing cluster topology of the network;
identify and optimize a data replication stream in use in the network;
determine a routing scheme based on the cluster topology; and
apply a predefined policy to at least one predefined router corresponding to a identified data replication stream.

11 . The system as claimed in claim 10 , wherein the one or more routers determines existing cluster topology of the network and comprises:
at least one edge router; and an edge router operating closest to the database server is a database node edge router.

12 . The system as claimed in claim 11 , wherein the database node edge router includes configured data comprising:
an Internet Protocol Address of the database node edge router; a Data base cluster unique identifier; and channel and associated policy configurations.

13 . The system as claimed in claim 11 wherein a master edge router optimizes data replication stream by enabling caching on edge router(s) and attaching a fingerprint to a payload associated with the cache.

14 . The system as claimed in claim 10 wherein the system is configured to:
create a path along with a label using a first router to route a data packet to a third router using cached data of a second router;
determine via the first router that a data packet to be transmitted to the third router is cached in the second router and using a cached fingerprint along with the path label;
receive by the second router the fingerprint with corresponding label, process the received packet data, convert the packet data with actual data and forward the actual data towards third router.

15 . The system as claimed in claim 10 , wherein a corresponding master DB node is configured to:
send REDO log information as a data packet to a subscriber node; intercept the packet data by a HIDER (master database edge router) to determine REDO log information; and cache the packet data after determining that it contains REDO log information and that cache is not already enabled.

16 . The system as claimed in claim 15 , wherein system is further configured to:
extract destination address of a data packet by the master database edge router via a lookup edge router information; generate a fingerprint for the packet data for a cache enabled edge router and store the fingerprint; label the data packet and forward the fingerprint; and generate a label by the master database edge router for the packet data and forward only the fingerprint when it is determined that a fingerprint already exists.

17 . The system as claimed in claim 10 wherein a master database edge router based on a determined topology of a DB cluster exhibits an optimal utilization of caching.

Assignees

• Huawei Tech Co Ltd

Inventors

• Ramamurthi Prasanna Venkatesh

Classifications

• H04L45/745

  Address table lookup; Address filtering · CPC title

• H04L45/02

  Topology update or discovery · CPC title

• G06F11/1471

  involving logging of persistent data for recovery · CPC title

• G06F17/30581 Primary

  Physics · mapped topic

• H04L67/1095

  Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes · CPC title