In this article, we are going to see the function and features of Database Availability Group in Exchange server which is to meet the high availability requirement.
Introduction:
Database Availability Group (DAG) is a feature that was introduced in Exchange 2010 for High Availability and Data Recovery for Mailbox Databases. It is referred as DAG in short, though DAG was introduced to provide High Availability for Mailbox Database, it also offers additional features for Transport High Availability by extending the transport boundary. We have discussed about this in Transport Dumpster article.
DAG:
DAG overwhelmed the High Availability model that was available in Exchange 2007, unlike 2007’s Local Continuous Replication (LCR) or Cluster Continuous Replication (CCR), DAG can be configured for multiple databases in different Geographical Locations. This flexibility in the architectural design provides ease of administration and managing complex environments. Single DAG can hold maximum 16 Mailbox Databases. DAG also eradicates traditional Fail-over Clustering, however Fail-over cluster role should be installed in the DAG server. It has in-built Fail-over Clustering which is called as Primary Active Manager. So this limits the dependencies on Windows Clustering. Separate Cluster logs will be created during the DAG configuration, hence fail-over is managed by Exchange itself. Fail-over and fail-back is faster than the past. Always the first created Mailbox server will be the Primary Active Manager.
How does DAG work?
Only servers that has a Mailbox role can be added in the DAG and every member server is capable of hosting both active and passive database copies. To learn more about DAG we should know the technical terminologies that are used.
Active Database Copy: It is the primary database of the mailbox with activation preference as 1. Always the active database hold it as 1. This priority is used by the Active Manager to decide on which copy should be activated at the time of Fail-over.- Passive Database Copy: Passive Database copies are replication of active copies. Each Passive copy will be placed in the other member servers of the DAG. The Database path and Logs folder Path should be identical as of the Primary server, else DAG will not function. Each copy will have an Activation Preference number, when the primary or Active Database copy fails, the Active Manager would immediately activates the best passive copy.
- Lagged Database Copy: Lagged Database copies are passive copies that are set with time delays. Generally the time delays can be of 2days to 7days, however as per the environment this can be changed. This is one way of preventing virus and other corruptions entering into this passive database (DB) i.e. If an Active DB becomes corrupted due to a virus, immediately it will be replicated to other passive DB, resulting corruption of that DB also. If there is a lagged copy, we stop the replication for that copy and activate that lagged copy and Replay logs again for the recent data.
- Log Shipping: As you all know that any email that arrives in the Exchange it will be converted as T-Logs, these T-Logs will be first played into the Active DB and gets stored in the Log folder. Now these new Logs will be shipped to the other Member servers, that holds this passive copy and again those T-Logs are replayed in the passive DB. This process is called Log Shipping.
Active Database Copy: It is the primary database of the mailbox with activation preference as 1. Always the active database hold it as 1. This priority is used by the Active Manager to decide on which copy should be activated at the time of Fail-over.So DAG literally uses Active and Passive Database copies and plays with. When the active copy goes down, immediately the best available passive copy will be activated so that users will not face in disruptions in accessing the emails. DAG is also used as Database backup and recovery. Exchange OST Recovery or mailbox recovery can also be done using a Recovery database.
Conclusion:
These are basics of DAG, and it is a very big topic which cannot be discussed or learned in single chapter, Lets discuss about troubleshooting DAG at the time of issues in another chapter.
Author Introduction:
Sophia Mao is a data recovery expert in DataNumen, Inc., which is the world leader in data recovery technologies, including repair pst data problem and word recovery software products. For more information visit www.datanumen.com
Leave a Reply