다음을 통해 공유

BizTalk Server: Disaster Recovery Strategies

  • 아티클

Introduction

As businesses around the world are becoming more global, applications are growing larger and larger. Thus, disaster recovery & business continuity have become necessity for the organizations.

In this article, we are trying to list down few techniques available in the market that can be used to set up the Disaster Recovery site and plan disaster recovery for BizTalk Server.

  1. BizTalk Server Log Shipping
  2. EMC Symmetrix Remote Data Facility (SRDF)
  3. EMC VPLEX with RecoveryPoint 

Please note that it is very important to understand each technique, both what is does and how it works? It is not at all recommended to directly implement these settings on your production environment and should be tested & evaluated properly in non-prod environments first.

 

Log Shipping

This is the Microsoft recommended and most commonly used DR strategy used across various organizations. This was introduced with BizTalk Server 2004 SP1 and was enhanced in BTS 2006.

What is Log Shipping and how it works?

http://msdn.microsoft.com/en-us/library/cc296856(v=bts.10).aspx

 

Configuring Log Shipping for BizTalk Server:

http://msdn.microsoft.com/en-us/library/gg608150.aspx

** **

EMC Symmetrix Remote Data Facility

In order to meet the initially said necessity, organizations are increasingly implementing solutions based on remote mirroring technologies. SRDF is one of such technique made available by EMC at storage level.

What is SRDF?

SRDF is basically mirroring on one disk device to another disk device in a separate Symmetrix system over a communication link. SRDF requires minimum of 2 Symmetrix disks (one the source and the other as target) in order to configure the data mirroring.

** **

How SRDF works?

SRDF offers following modes of operations & configuration options for specifying how data transfer is performed between the source & target disks.

  1. Synchronous Mode
  2. Semi-Synchronous Mode
  3. Adaptive Copy Mode.

** **

Synchronous Mode:

SRDF is configured in synchronous mode for those cases where data must be written to both source & target system before the application is to continue working. Below are the sequences of events that occur as part of synchronous mode:

  1. Data is written to source Symmetrix disk cache.
  2. This data is then written to target Symmetrix disk cache.
  3. The target Symmetrix disk then acknowledges the receipt of data.
  4. The I/O complete signal is then sent back to application.

As clear from the above description, data is written to both the Symmetrix disks (local & remote) before I/O complete signal is sent back to application, this impacts the application performance & introduces a delay while processing huge data volume. So, in order to improve the performance, huge data activity should be spread over as many logical volumes as possible so as to reduce the overall I/O rate per volume. This is known as Host Level Stripping.

 

Semi-Synchronous Mode:

Semi-synchronous mode first writes the data to the local Symmetrix disk, completes the I/O operation and then synchronizes the data with the remote disk. Since the I/O is completed before synchronizing the data with the remote disk, this mode provides an added performance advantage. A second write is not accepted by the local disk until the remote disk is synchronized, the application can continue performing read operations though.

Below are the sequences of events that occur as part of semi-synchronous mode:

  1. Data is written to source Symmetrix disk cache.
  2. The I/O complete signal is then sent back to application.
  3. This data is then written to target Symmetrix disk cache.
  4. The target Symmetrix disk then acknowledges the receipt of data.

 

Adaptive Copy Mode:

In this SRDF mode, data is transferred from the source disk to the remote disk without waiting for

Acknowledgment and allows a lag between source and target disks of a specified number of I/Os as selected by user. This operational mode is not recommended for use when mirroring for disaster recovery.

EMC SRDF Product Description Guide

http://www.emc.com/collateral/software/symmetrix-remote-data-facility.pdf

 

**EMC VPLEX with RecoveryPoint **

 

What is VPLEX?

VPLEX is basically a storage virtualization device that sits between the storage array & the hosts and virtualizes the presentation of storage arrays, including non-EMC arrays as well. VPLEX storage appliance: 

  1. Presents the storage array to VPLEX, instead of Host/Servers.
  2. Configures the storage within VPLEX virtual storage arrays.
  3. Then presents this VPLEX virtual storage arrays to Hosts/Servers.

RecoveryPoint, as defined on Wiki, is a continuous data protection solution offered by EMC Corporation which supports asynchronous and synchronous data replication of block-based storage. RecoveryPoint Continuous Remote Replication (CRR) enables a replica in a remote site.

Different VPLEX Configurations

http://www.emc.com/storage/vplex/vplex.htm#!details

How VPLEX Works?

A combination of VPLEX and RecoveryPoint configured together, results in Zero RTO (Recovery Time Objective) & RPO (Recovery Point Objective). 

Configure EMC VPLEX along with EMC RecoveryPoint to set up DR

http://www.emc.com/collateral/hardware/white-papers/h10746-vplex-recoverpoint-oracle-dr-wp.pdf

EMC RecoveryPoint

http://www.emc.com/storage/recoverpoint/recoverpoint.htm

Advantages & Disadvantages of EMC VPLEX

http://emcsan.wordpress.com/2014/02/11/what-is-vplex/

** **

Conclusion

One should evaluate & choose the right DR strategy (or a combinatrion) in order to achieve minimum Recovery Time Objective (RTO) and Recovery Point Objective (RPO)

See Also

Another important place to find an extensive amount of BizTalk related articles is the TechNet Wiki itself. The best entry point isBizTalk Server Resources on the TechNet Wiki.