1

I am deploying redash in Openshift and the single instance mode works well.
But I would like to make it multi instance and highly available so that it can overcome pod removals which can be caused by node draining etc.
Has any one done a HA version of redash?

1 Like
2

I’m also looking for HA redash.

3

any updates? were you able to get it working?

4

I tried it last year.
1.Configure multiple docker-compose.yml configuration files to start multiple Redash instances on different servers or on different ports on the same server
2.By change the config of nginx, then nginx service load-balanced and forwarded all requests to different instances of the Redash service.

But there are two problems I haven’t solved：

  1. When A query is being executed on instance A, the client will continue to send requests to the nginx service to get the result of the query. If the request is forwarded to instance B by nginx, the client will display “query in queue”（It may be confusing for users）.
    2.I’ve only tested the PG database cluster, but it hasn’t actually been used in production. High availability may not be dependent on a single database instance。