dapr
Multiple servers, different placements and schedulers, if you want to share data
Multiple servers with different placements and schedulers, if you want to share data, is it enough for me to deploy etcd separately and connect them to the same etcd? How can actors ensure the same concurrency? If I connect to different placements, will he still ensure that they are the same concurrent?
@bbhxwl I think you are asking about whether Dapr Scheduler & Placement support multi-cluster (?)- no Dapr doesn't support this and should be the case that each Scheduler cluster connects to it's own etcd database. For this to work, it would require a single Scheduler set, a flat network, and a shared trust anchor/domain across the two clusters.
@bbhxwl I think you are asking about whether Dapr Scheduler & Placement support multi-cluster (?)- no Dapr doesn't support this and should be the case that each Scheduler cluster connects to it's own etcd database. For this to work, it would require a single Scheduler set, a flat network, and a shared trust anchor/domain across the two clusters.
Does a single daprd need to connect to a separate Scheduler and Placement, or is it recommended to connect multiple daprds to the same Scheduler and Placement? I am confused about this aspect, and currently I feel that only the state of dapr is the easiest to use.
Multiple daprds connect to the same control plane, which include scheduler and placement.
https://docs.dapr.io/concepts/dapr-services/
Multiple daprds connect to the same control plane, which include scheduler and placement.
https://docs.dapr.io/concepts/dapr-services/
Isn't that what I said? As long as etcd is deployed separately, the scheduler can connect to the etcd cluster for data sharing?
Multiple daprds connect to the same control plane, which include scheduler and placement.
https://docs.dapr.io/concepts/dapr-services/
For example, if three schedulers are connected to the same etcd and three daprds are connected to three different schedulers, isn't it the same?
Daprd connects to all scheduler replicas
https://docs.dapr.io/concepts/dapr-services/scheduler/