High availability and site resilience
The high availability model of the mailbox component has not changed significantly since Exchange 2010. The unit of high availability is still the database availability group (DAG). The DAG still uses Windows Server fail-over clustering. Continuous replication still supports both file mode and block mode replication. However, there have been some improvements. Fail-over times have been improved to reduction as a result of transaction log code enhancements and deeper checkpoint on the passive databases. The Exchange Store service has been re-scripted in managed code. Now, each database executes under its own process that isolates store issues to a single database.
Exchange 2016 uses DAGs and mailbox database replicas, along with other functional features such as retention policies, single item recovery, and lagged database copies, to offer high availability, site resilience, and Exchange native data security and protection. The high accessibility and availability platform, the Exchange Information Store and the Extensible Storage Engine (ESE), have all been improved to offer greater availability, simple and easier management, and to reduce costs. These improvements include:
- Managed and control availability With managed availability, internal monitoring and recovery-oriented functional features are tightly coupled and integrated to help prevent failures, proactively restore services, and initiate server failovers by default automatically or alert administrators to take measures. The objective is on monitoring and managing the end user exposure, experience instead of just server and component uptime to help keep intact the service continuously available.
- Managed Store Refer the Managed Store section.
- Support for several databases per disk Exchange 2016 consists of improvements that enable you to support several databases (amalgamations of active and passive copies) present on the same disk, thereby benefiting larger disks in terms of capacity and IOPS as effective efficiently as possible.
- Automatic reseed Empowers you to faster restore database redundancy after disk failure. In case a disk fails, the database replica stored on that disk is replicated from the active database copy to a spare unoccupied disk on the same server. In case multiple database copies were stored on the failed disk, then they can all be automatically re-seeded on a spare unused disk. This enables quicker reseeds, as the active databases are expected to be on multiple servers and the data is replicated in parallel.
- Automatic recovery default from storage failures This functional feature continues the innovation which was introduced in Exchange 2010 to permit the system to recover from failures which affect resiliency or redundancy. In addition to the Exchange 2010 bugcheck behaviors, Exchange 2016 consists of additional recovery characteristics for long I/O times, excessive memory utilization or consumption by MSExchangeRepl.exe, and severe cases where the system is in such a state that threads can’t be scheduled.
- Lagged copy improvements Lagged copies can now utilize automatic log play down to care for themselves (up to a certain extent). Lagged replicas will automatically play down log files in various categories of situations, such as single page restore and low disk space scenarios. In case the system detects that page patching is needed for a lagged copy, the logs will be automatically replayed into the lagged replica. Lagged copies will also activate to invoke this auto replay functional feature when a low disk space threshold has been reached, and when the lagged copy has been found as the only available copy for a specific period of time. In addition, lagged copies can advantage to Safety Net, making recovery or activation much simpler and easier. Safety Net is enhanced functionality in Exchange 2016 based on the transport dumpster of Exchange 2010.
- Single copy alert improvements The single replica alert which was introduced in Exchange 2010 is no longer a segregated scheduled script. It’s now coupled and integrated into the managed availability components within the system and is a native function within Exchange.
- DAG network auto-configuration DAGs networks can be automatically configured by the system depending on configuration settings. In addition to manual configuration choices, DAGs can also distinguish among MAPI and Replication networks and configure DAG networks automatically.
In Exchange 2016, Managed Store is the synonym of the Information Store processes,
Microsoft.Exchange.Store.Service.exe and Microsoft.Exchange.Store.Worker.exe.
The new novice Managed Store is scripted in C# and is tightly coupled and integrated along with the Microsoft Exchange Replication service (MSExchangeRepl.exe) to offer higher availability through enhanced resiliency.
In addition, the Managed Store permits more granular management and control of resource consumption, and has enhanced diagnostics for quicker root cause analysis.
The Managed and control Store works with the Microsoft Exchange Replication service to control and manage mailbox databases that continue to use the Extensible Storage Engine (ESE) database engine. Exchange 2016 consists of significant changes to the mailbox database schema which provide many optimizations over earlier versions of Exchange. It is the responsibility of the Microsoft Exchange Replication service for all service availability concerned to Mailbox servers. These architectural amendments enable faster database failover and robust physical disk failure handling comparatively.
The Managed Store consumes the same search platform as SharePoint Server 2016 to offer more robust indexing and searching in comparison to Microsoft Search engine which was used in previous versions of Exchange.