DexPro
;Dynamic Seg- mentation
Data
Centrali-
sation Manager
De-
duplicator
DexPro
Plan Manager
Email
Marketing
System
Integra-
tion Server
The
Replicator
Distributed Database Availability (DDA) and WAN Traffic
Reduction
FEATURES
Send and Receive | Will send and receive data and database schema changes to participating databases |
SQL Server 2005 compatible | Will run on SQL Server 2005 |
SQL Server 2008 compatible | Will run on SQL Server 2008 |
No Structural changes to tables | Does not involve forcible addition of new table columns to accommodate Replication ID |
No database size constraints | No maximum number of columns or rows size in replicated tables |
Data and Schema Replication | Replication of all changes in data and database schema as they occur |
Low latency | No waiting time between reading from principal and writing to replica |
Corruption Replication Protection | Error log entries are not replicated thus eliminating possibility of passing application corruptions |
Conflicting URN Protection | Eliminates possibility that the same URN is allocated to different records from different systems |
Unlimited databases and sites | There is no limit to the number of servers and databases that can be replicated by MMM |
No distance limitations | Servers involved in replication can be any distance apart |
Site failure protection | Provides protection from database, disc, server, connection or any other site failure |
False fail over protection | Allows administrator to configure when to treat interruption as fail over event |
Automatic Fail over | Switches users to alternative server in the event of failure (subject to connection availability |
Automatic Fail back | Switches users back to failed server when recovered |
Automatic Redo | Automatic rewind to last backup recovered and redo ensuring no loss of data |
Front-end independent | No need to modify front-end applications for automatic Fail over and Fail back |
No need for fail over witness | Does not require witness to cover exposed mirror during fail over |
No need for Distributor | Does not require Publisher, Distributor, Subscriber(s) to perform Replication |
Asynchronous Mirroring mode | Does not wait for commit from replica thus saving 5-10% overhead in LAN and more in WAN |
Mode independent fail over | No need to be in synchronous mode for fail over |
No naming constraints | No need for replicated databases to have the same name |
Replicas are in read/write state | Replicated databases are in normal state and can be used for SELECT statements |
Bulk-logged record replication | Replication of bulk-logged operations |
No Recovery Model constraints | Principal database does not have to be in Full Recovery Model to replicate |
No need for manual monitoring | No need for DBA to monitor transaction log and log file synchronization |
HCL Independent | Does nor require use of components from Microsoft Hardware Compatibility List |
No silent data corruption | Logical groupings of database columns or rows remain intact during replication |
:: : TOP ::
Multi-directional Mirroring and Replication - Step 1: Multi-Directional
Updates
:: TOP ::
Step 2: Auto-Switching
:: TOP ::
Step 3: When DB Server 3 is up again
:: TOP ::