-
Hi there!
First of all, it works like a charm when partitions are empty. Not bad indeed! :)
When partitions are fulfilled, it's hardly ever possible to start the cluster in general. I keep getting…
-
I have to say that it is currently for me quite hard to debug/follow our current log especially on chaos testing.
We should try to improve our logging, first of all in our chaos tests we use the d…
-
**Describe the bug**
Lately we observed in our long running test suite that the processing of one partition (eventually most of the partitions) stop processing.
![processing-stopped](https://use…
-
I would like to discuss an alternative approach of our DistributedLog primitive.
Please share your thoughts about this.
### Idea
We have our custom primitive, but instead of writing the events…
-
**Describe the bug**
`Expected to execute side effects for event 'LoggedEvent [type=0, version=0, streamId=3, position=103495829491608, key=-1, timestamp=1574238760521, sourceEventPosition=-1]' suc…
-
**Describe the bug**
In our long running test suite and also in our chaos cluster I saw that one pod was killed by k8s and this cause to break the cluster.
### Long running
Setup:
* 8 Part…
-
**Describe the bug**
When new partition is installed and no exporter are configured then the ExporterState is cleaned up. It can happen that concurrently to that the partition is closed again, beca…
-
**BREAKING CHANGE**
**Is your feature request related to a problem? Please describe.**
While we already use Raft as the back-bone for our consistency guarantees, but quite indirectly at the mome…
-
**Describe the bug**
Yesterday I setup a new cluster for chaos testing and choosed preemptable nodes as a setting.
In the morning I started the cluster and did some test everything worked fine, bu…
-
```
root@XXXXX:/bookkeeper# bin/dlog admin unbind -f distributedlog://XXXX:2181/XXXXX
Fail to run tool org.apache.distributedlog.admin.DistributedLogAdmin :
java.lang.IllegalArgumentException: Zoo…