gyang94 commented on code in PR #2757:
URL: https://github.com/apache/fluss/pull/2757#discussion_r2901195990
##########
fluss-common/src/main/java/org/apache/fluss/config/ConfigOptions.java:
##########
@@ -96,8 +96,65 @@ public class ConfigOptions {
.stringType()
.noDefaultValue()
.withDescription(
- "The directory used for storing the kv snapshot
data files and remote log for log tiered storage "
- + " in a Fluss supported filesystem.");
+ "The directory in a Fluss supported filesystem for
remote data storage. "
+ + "This configuration is required. "
Review Comment:
This configuration is required now? suppose a fluss server should be able to
run without a remote storage.
OK... see the comments above.
> As I updated in the document. We should always set remote.data.dir, as the
recently introduced producer offsets and kv snapshot lease files are not belong
to a specific table.
> kv snapshot lease dir:
{$remote.data.dir}/lease/kv-snapshot/{leaseId}/{tableId}/
> producer offset dir: {$remote.data.dir}/producers
> I think we must keep remote.data.dir to store producer offsets and kv
snapshot lease files for now.
These features rely on the remote storage.
My view is: generally we should allow the server starts when no remote
storage configured. When users case need these two features, there will be an
exception to tell users that they need to setup remote storage before move on.
What do you think?
--
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]
