pandaapo commented on code in PR #4544:
URL: https://github.com/apache/eventmesh/pull/4544#discussion_r1387813255
##########
eventmesh-runtime/build.gradle:
##########
@@ -72,6 +72,9 @@ dependencies {
implementation project(":eventmesh-webhook:eventmesh-webhook-api")
implementation project(":eventmesh-webhook:eventmesh-webhook-receive")
+ implementation project(":eventmesh-retry:eventmesh-retry-api")
+ implementation project(":eventmesh-retry:eventmesh-retry-rocketmq")
Review Comment:
> This is because the runtime needs to be loaded into a specific middleware
SPI implementation
When using standalone mode, even if `retry.storageEnabled` is configured to
true, we can directly use the retry in memory provided by EventMesh without the
need to load the specific implementation of SPI. Of course, at this point, a
log can be provided to prompt the user that `retry.storageEnabled` does not
work because they are using standalone mode.
In addition, when it is not standalone and `retry. storageEnabled` is
configured to true, if the specific implementation of SPI cannot be loaded, the
user will be prompted by printing a log, and then we use the retry in memory
provided by EventMesh as a backup.
This can achieve more thorough decoupling, wouldn't it be better?
当使用standalone模式时，即使`retry.storageEnabled`配置为true，就可以直接采用EventMesh提供的内存中重试，而不必加载SPI的具体实现。当然这时可以给出日志提示用户因为使用standalone模式，所以`retry.storageEnabled`不起作用。
另外还可以在非standalone且`retry.storageEnabled`配置为true时，如果加载不到SPI的具体实现，打印日志提示用户以后，就采用EventMesh提供的内存中重试作为一种兜底。
这样可以做到更彻底的解耦，不是更好吗？
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
