Repository navigation
CASSSIDECAR-492: CDC log_only mode should not require Kafka producer … - #389
Conversation
| cdcConfigs(); | ||
| return cdcConfigAccessor.isAvailable() | ||
| && !kafkaConfigMappings.isEmpty() | ||
| && (logOnly() || !kafkaConfigMappings.isEmpty()) |
There was a problem hiding this comment.
This starts to feel a little bit convoluted (not this PR's fault).
CEP-44 offers Cdc over kafka, but, kafka is just a data sink we are using for CDC. Logs could be just another one. What do you think about a broader refactor to support kafka, logs, or any other type of "sending CDC events"?, that could even work in parallel?
For example:
- Log AND kafka
- Log AND AWS SNS
Having said that, I understand the scope of this proposal may be a bit bigger, and I guess it'd be ok to have it as part of another Jira, but, what do folks think?
There was a problem hiding this comment.
Let's leave the refactoring in a dedicated JIRA.
Sharing my thoughts — see if they make sense.
Log is not a data sink. It is a by-product of the process. The patch is to introduce a mode to avoid publishing to data sink, which is kafka specifically as of today.
If we want to introduce other data sink options, (yet to be discussed), the selection would be among the data sinks. Log is configured separately.
|
|
||
| if (conf.logOnly()) | ||
| { | ||
| LOGGER.warn("CDC log_only is set; mutations will be logged and not published to Kafka"); |
There was a problem hiding this comment.
Shouldn't this be an info? Configuration has been explicitly selected for this.
jyothsnakonisa
left a comment
jyothsnakonisa
left a comment
There was a problem hiding this comment.
Looks good to me, can u rebase the PR and also run CI using circle CI and paste the link
|
+1 |
d6b43b7 to
7054cfa
Compare
7054cfa to
4105b04
Compare
4 participants
…configuration
Patch by Shruti Sekaran; Reviewed by TBD for CASSSIDECAR-492