Details
-
Improvement
-
Status: Closed
-
Minor
-
Resolution: Fixed
-
None
-
None
-
None
Description
- In config json, accept multiple zookeeper to tolerate single point of failure
- In config json, partition number can be omitted, read it from zookeeper. Or another option is keep the configure, and validate it against zookeeper at runtime. The latter makes the mapping between partition and sharding clearer.
- Should report the data up-to-date time in some way. On a second thought, I don’t think segment end time is the right place, but some other API.
Attachments
1.
|
accept multiple zookeeper in KafkaConfig | Closed | Qianhao Zhou | |
2.
|
read topic partitions from Kafka instead of metadata | Closed | Unassigned | |
3.
|
new api for streaming to report up-to-date time | Closed | Unassigned |