SQL Query hangs when one node if suspect or dead

I am just starting with cockroachDB. It seems to be a great and simple solution.
I have setted up an insecure node cluster in a my local network, to run some tests.
Everything seems to work. But as soons as I put one node down (stop or unplug the server), the sql querying hangs. The query answer results only as soon as I put all the missing nodes online.

Strangely I get a response when I query another DB (the BankDB from tutorial examples apps https://www.cockroachlabs.com/docs/stable/hello-world-example-apps.html) but all the other DB seems to stay down, until I restart the missing nodes

Am I missing something ?


Is this happening on a Linux OS? I have the same issue on macOS, but not on Linux.

Hey Boris.
Happened with 3 machines running Windows. I will try on Linux as soon as I have enough Linux workstation running this system and some time.


It is the case that when a node dies, we need to wait some time for failure detection. In the case of normal writing transactions this should be on the order of seconds. In the case of schema changes, this will be around 5 minutes. What sort of queries are you using and what sort of hanging are you seeing?

on macos: just normal select * from table1 limit 1;
same query on linux, but no issues.
on macos, the dashboard also stops working when one out of three instances is killed.