Hi there I have deployed dagster and daemon in 2 clusters to

Question

Hi there I have deployed dagster and daemon in 2 clusters to be on the safe side in production However this is causing two behaviours I was not expecting 1 schedules and sensors trigger 2 runs I assume one for each cluster How can I avoid this Do I have to query GraphQL or is there a better way 2 I am getting this log message gt Another SENSOR daemon is still sending heartbeats You likely have multiple daemon processes running at once which is not supported Last heartbeat daemon id 64d45a3b a6b8 481f a036 dd4901c23182 Current daemon id 0101ad4d 983b 4d48 a09a 9bdcd46dd26a

johann · Answer

Hi Bernardo we don t currently support running simultaneous daemons

johann · Answer

Distributing schedules and sensors across multiple clusters is possible in Cloud thanks to the agent model we used there it s likely that we ll bring that to OSS in the future

Bernardo Cortez · Answer

But what do you expect to happen with this I am getting a CodeUnreachable error and I was wondering if this could be related to it

Bernardo Cortez · Answer

in Pre production everything is working apart from the fact that the duplicated daemons trigger duplicated runs But in Production sensors break with CodeUnreachable Is it related to the multiple daemons

johann · Answer

I m not sure Having multiple daemons sets you up for undefined behavior but not necessarily CodeUnreachable

johann · Answer

That error means the daemon can t connect to a grpc server

Bernardo Cortez · Answer

so there is a higher probability of being caused by something else

johann · Answer

I would see if you still hit the error with one Daemon first