Paweł Gontarz
Paweł Gontarz
Hello @ddddddcf , today it does not work because chproxy uses its own generated `query_id` to kill queries exceeding timeouts. > ... the generated query_id is used for query logging...
For future improvement to avoid this kind of complicated issues we should have a fake server handler scoped to each test case. Tests will be a bit longer, but I...
Hi @mr-karan , this feature has been released along with the version `v0.17.0`
Hello @bzon, I agree with your proposal. Would you mind submitting a PR?
Hello @OmarBazaraa, no it is not. You'd have to create 20 users.
Hello @kasnet , do you mean load distribution based on weight? Such as having topology with 3 Clickhouse nodes, you'd want 50 % of load on node 1 and 25...
Fix for that has been released following the commit #146
I don't think we want to go that way. If you have some specific needs we should integrate it within github actions. Probably no longer relevant cc @sigua-cs
Hello @pavelnemirovsky , github actions are now available in chproxy. Could you rebase your PR so we can leverage the fix you'd submitted for session timeout?
Hello @rlazimi-dev, we haven't planned this in near future. I'm not sure how would you like it to work. Feel free to propose a design and use cases envisioned