You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am using the simple scalable helm chart of Loki and have two nodes in my monitoring kubernetes cluster.
So far so good, this config is working fine, from the official grafana docs:
Now I was trying test if i everything stays up and running if i kill a node in the cluster, so I did and I kept getting some error along these lines: "too many unhealthy instances in the ring". After reading into the replication factor I realized it probably happens because I am running only 2 instead of 3 replicas each and the default factor is 3.
I have tried setting replication_factor=1 but that seems to still need 2 active replicas for a successful write. According to the docs the factor results from ( replication_factor / 2 ) + 1.
A factor of 0 is not allowed.
My question is: is there a way of making my setup work?
I appreciate any input.
Thanks
The text was updated successfully, but these errors were encountered:
I have now also tried using the deploymentMode: SingleBinary and replicas: 2 for it. It also does not work with either settings for replication_factor, the default 3 nor 1.
Is there no way of running Loki in a highly available matter with just 2 replicas?
Hi everyone,
I am using the simple scalable helm chart of Loki and have two nodes in my monitoring kubernetes cluster.
So far so good, this config is working fine, from the official grafana docs:
Now I was trying test if i everything stays up and running if i kill a node in the cluster, so I did and I kept getting some error along these lines: "too many unhealthy instances in the ring". After reading into the replication factor I realized it probably happens because I am running only 2 instead of 3 replicas each and the default factor is 3.
I have tried setting
replication_factor=1
but that seems to still need 2 active replicas for a successful write. According to the docs the factor results from( replication_factor / 2 ) + 1
.A factor of 0 is not allowed.
My question is: is there a way of making my setup work?
I appreciate any input.
Thanks
The text was updated successfully, but these errors were encountered: