unable to run Mimir in single ingestor - too many unhealthy instances in the ring

[kirankumar-grootan]

unable to run Mimir in single replica ingester via helm chart

mimir:
  structuredConfig:
    store_gateway:
      sharding_ring:
        replication_factor: 1
    ingester:
      ring:
        replication_factor: 1
  ingester:
    replicas: 1

when I run query from grafana it shows too many unhealthy instances in the ring. I dont need HA for dev cluster thats why Im reducing the ingester to 1

[aknuds1]

Converting to discussion, as there's no evidence of a bug.

[kreeger]

I too had this issue, and played around with my Helm values until I got it working. I was originally missing replication_factor: 1 in my structuredConfig! You'll also need zone_awareness_enabled: false, since to disable replication on ingested, and store_gateway means disabling zone awareness.

This is what did the trick for me (after setting zoneAwareReplication: { enabled: false } on store_gateway and ingester).

mimir:
  structuredConfig:
    ingester:
      ring:
        zone_awareness_enabled: false
        replication_factor: 1
    store_gateway:
      sharding_ring:
        zone_awareness_enabled: false
        replication_factor: 1

[jamestjw]

if you are using the grafana helm chart, it isn't necessary to set zone_awareness_enabled. What I didn't realise I had to do was that I had to first delete the entire mimir deployment before reducing the number of replicas.