OpenEdge Replication startup and initialization process
- Last Updated: March 30, 2020
- 2 minute read
- OpenEdge
- Version 12.2
- Documentation
During the OpenEdge Replication startup and initialization process,
the OpenEdge Replication server attempts to contact the OpenEdge Replication agent through
the target database broker on the port specified in the control-agent section
of the source database replication properties file. Once the OpenEdge Replication server
makes contact with the OpenEdge Replication agent, a handshaking process takes
place.
During the handshaking process, OpenEdge Replication:
- Determines if the source and target databases are identical
- Verifies that the target database was created from the source database
- Performs synchronization
- Allows database connections
On the replication source database, connections are allowed before the handshaking process, regardless of the setting of the defer-agent-startup property. Source update activity, however, is halted until the replication server is granted a shared schema lock. Once the replication server gets the shared schema lock, source database client update activity resumes.
On the replication target database, read only clients are not allowed until startup synchronization has completed unless the previous replication agent session was running under a replication plus license. In that case, read only clients are allowed during startup synchronization.
If the defer-agent-startup property is set
to a valid non-zero value, source database activity is allowed once the source database is
started. Source database activity does not halt until the very end of the synchronization
process, when the OpenEdge Replication server completes the synchronization process. When
synchronization is complete, the Replication server reinserts itself back into the AI Block
write process, where the OpenEdge database is unblocked, and normal database and OpenEdge
Replication activity continues.
On the replication source database, connections are allowed before the handshaking process, regardless of the setting of the defer-agent-startup property. Source update activity, however, is halted until the replication server is granted a shared schema lock. Once the replication server gets the shared schema lock, source database client update activity resumes.
Before synchronization begins, the OpenEdge Replication server attempts to
connect to its configured agent(s). If the server cannot connect to its configured agent(s) on
the first connection attempt, it goes into deferred agent startup. While in the deferred agent
startup state, the OpenEdge Replication server waits five minutes, and then attempts the
OpenEdge Replication agent connection again. The OpenEdge Replication server remains in this
state until all agent(s) are connected, or until the time-out specified in the defer-agent-startup property is reached.
Startup, initialization, and synchronization are performed once
all the agents are connected. If you do not want to wait for the
five-minute intervals between connection attempts to the agent(s),
you can force agent connection using the DSRUTIL function connectagent.
Schema updates are not allowed while the OpenEdge Replication server is performing synchronization. If schema updates are being performed when failure recovery synchronization begins, source database updates block until failure recovery is complete.