[destination-clickhouse] Clickhouse cluster support #36330
rsitrifork
started this conversation in
Connector Ideas and Features
Replies: 1 comment
-
+1 on this! Connections to ClickHouse client managed clusters are always made to one of the nodes. Basically all SQL statements sent to that node needs to have the “on cluster”-clause to ensure that databases and tables are propagated to all nodes. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Connector Name
destination-clickhouse
Connector Version
1.0.0
What step the error happened?
Configuring a new connector
Relevant information
Clickhouse is installed on AKS as a cluster with 3 shards and 2 replicas. Airbyte is installed in the same cluster.
The "airbyte_internal" table that is created is lacking the "on cluster default" specifier and as such is only created on the current replica. When connecting to clickhouse using a loadbalancer (in my case a kubernetes service) this means that sometimes the database does not exist.
Relevant log output
No response
Contribute
Beta Was this translation helpful? Give feedback.
All reactions