diff options
author | Harald Musum <musum@verizonmedia.com> | 2021-02-08 14:46:32 +0100 |
---|---|---|
committer | Harald Musum <musum@verizonmedia.com> | 2021-02-08 14:46:32 +0100 |
commit | 3bf63b5abac30185ff0dd38b6ed89cba9b16bb4b (patch) | |
tree | cfb5381380283a069550c2f3004e915f66aed7bb | |
parent | 21277de10864da6edd003c4537f1c069fa3501d1 (diff) |
Allow only one node down at a time for cluster controller clusters
If we are replacing nodes and have 3 old and 3 new nodes in a cluster,
allowing 50 per cent down would lead to 3 being allowed to go down.
When deploying to remove the 3 old nodes there might be (for a short time)
config that says that there should be 6 nodes in the zookeeper cluster.
This mean 4 of them need to be up for the cluster to have a quorum.
This will not work in this case.
-rw-r--r-- | orchestrator/src/main/java/com/yahoo/vespa/orchestrator/policy/HostedVespaClusterPolicy.java | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/orchestrator/src/main/java/com/yahoo/vespa/orchestrator/policy/HostedVespaClusterPolicy.java b/orchestrator/src/main/java/com/yahoo/vespa/orchestrator/policy/HostedVespaClusterPolicy.java index d3e7afba25f..3088fd7ae0a 100644 --- a/orchestrator/src/main/java/com/yahoo/vespa/orchestrator/policy/HostedVespaClusterPolicy.java +++ b/orchestrator/src/main/java/com/yahoo/vespa/orchestrator/policy/HostedVespaClusterPolicy.java @@ -165,7 +165,7 @@ public class HostedVespaClusterPolicy implements ClusterPolicy { if (ServiceType.CLUSTER_CONTROLLER.equals(clusterApi.serviceType())) { // All nodes have all state and we need to be able to remove the half that are retired on cluster migration - return ConcurrentSuspensionLimitForCluster.FIFTY_PERCENT; + return ConcurrentSuspensionLimitForCluster.ONE_NODE; } if (ServiceType.METRICS_PROXY.equals(clusterApi.serviceType())) { |