You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently any queue configured for error, audit, servicecontrol or servicecontrol monitoring skips routing. It would be very convenient instance routing would be applied to any of these queues.
Customer only has to maintain a single instance mapping file
Customer can have round-robin distribution between multiple instance for a queue
Other transports do use routing mapping when sending message like for example sqlt.
Possible behavior:
If a instance is specified on the address, use it.
If no instance is specified, use the instance resolved from the instance mapping, if no instance is round throw exception that no instance could be resolved indicating to add this to the instance mapping source.
Example:
Customer wants to scale-out a current one machine deployment and now needs specify an instance in any of the API's while the instances are already managed in the instance mapping file.
The text was updated successfully, but these errors were encountered:
ramonsmits
changed the title
Consider adding to use instance mapping to route message to another host/instance
Routing: Consider adding to use instance mapping to route message to another host/instance
Jun 10, 2020
Research is needed to understand where the translation from logical to physical can happen might we use the current API's to contain a logical endpoint.
The plugins use the low-level dispatch API's. If we want to map logical to physical this would likely require changes in these plugins.
We could alternatively go for a physical to physical translation.
Currently any queue configured for error, audit, servicecontrol or servicecontrol monitoring skips routing. It would be very convenient instance routing would be applied to any of these queues.
Other transports do use routing mapping when sending message like for example sqlt.
Possible behavior:
Example:
Customer wants to scale-out a current one machine deployment and now needs specify an instance in any of the API's while the instances are already managed in the instance mapping file.
The text was updated successfully, but these errors were encountered: