Roy Dahan comments

Results 453 comments of


                                            Roy Dahan

full_storage_utilization_test: reclaim and scale out tests

One more thing, I don't understand the motivation of using performance_regression tests here. Is the intention of the test to measure the latency and duration? It's need to be written...

`GrowShrinkClusterNemesis` fails when current number of nodes is less the initial number of nodes

I don't understand what was the issue here. The terminate nemesis node couldn't remove the node? It shouldn't raise a critical.

`GrowShrinkClusterNemesis` fails when current number of nodes is less the initial number of nodes

ok, now I understand. It's fine to fail it as critical.

fix(disrupt_terminate_and_replace_node): raise critical event on failure

Why is it needed?

Session not reconnected after rolling upgrade

Should it reconnect? How should it happen. In any case, I don't think system_auth with Rf=1 is a relevant case.

3.x: Add RandomTwoChoice load balancing policy

I think the PR got neglected for two reasons: 1.The write workloads weren't good enough and it was not followed up. 2. Probably there is no much demand for it?...

Ensure scylla-rust-driver handles zero-token nodes properly

We can use ccm now to test if it works and at least know if something need to be done here.

driver reports connection errors for decommissioned nodes and a delay for cassandra-stress, resulted with: "Command did not complete within 12870 seconds!"

Are you sure it's not related to this one: https://github.com/scylladb/scylla-cluster-tests/issues/8855#issuecomment-2614500840

Bump io.debezium:debezium-core from 1.4.1.Final to 2.3.0.Final

Please prioritize.

Bump io.debezium:debezium-core from 1.4.1.Final to 2.3.0.Final

You can also try the SCT case that tests connector. (I don't know how you pass a non official release of the connector)