When you click Settings in the Admin Console, the Cluster Settings dialog box appears, enabling you to make runtime changes to several configuration parameters in your cluster. When you save your settings for the first time, the Cluster Settings UUID field displays the universal unique identifier (UUID) of the anchor stream that contains the cluster settings. All changes will persist in the cluster across reboots.
Note: Do not set cluster-wide persisted parameters in the individual node configuration files because these values must be the identical for the entire cluster. Any values specified in a node configuration file are overwritten by the cluster settings in the Admin Console.
About Cluster Settings
The cluster settings are now automatically propagated to all nodes in the cluster. Adding the cluster settings UUID to your node or cluster configuration file is not required.
The first time you change a cluster-wide setting, the cluster settings UUID is automatically generated and propagated to all nodes in the cluster.
Best Practices for Maintaining Cluster Settings
For best results, use the following guidelines:
- Set the
cluster.nameconfiguration parameter value before you change any cluster settings. After the cluster settings UUID is set, do not change the value.
- If you set the
cluster.nameand change it or have no
cluster.nameset and add one after you have a cluster settings UUID, you may encounter conflicts in your settings UUID. Contact Support for instructions on resetting your cluster settings UUID.
- If you move a volume between clusters, the volume will keep its original cluster settings and
cluster.nameparameter value. To overwrite the original settings with the new cluster settings, you must insert the volume into its new location while the remaining nodes are running. If all volumes are mounted at the same time, the settings from the originating cluster will be detected first and become "master" in the new cluster.
You can update the logging Host, Port, and Level options to temporarily or permanently redirect the cluster logs to a different location or log level. This can be useful when troubleshooting an issue in the cluster that requires a more granular report of cluster activity.
In addition, you can check the Audit Logging checkbox to send audit level events to the syslog, independent of the log level.
This option allows you to set the frequency of multicast broadcasts within the cluster. The higher the percentage you enter in the Multicast % field, the more frequently the cluster will communicate UUIDs to the cluster nodes and other services, such as Content Router.
Netmail Store will automatically initiate recovery of replicas and erasure-coded segments known to be on a cluster volume that is no longer present in the cluster. Two recovery processes, failed volume recovery (FVR) and erasure coding recovery (ECR), are started for every volume detected as missing. Netmail Store will announce both when it starts each process as well as when each completes, which may be relatively quick if there are no replicas or segments to recover.
The Suspend option allows you to temporarily suspend both volume recovery processes in the cluster for situations where data is not actually at risk, but a network or power outages have taken one or more nodes (or an entire sub-cluster) offline. Suspending volume recovery by selecting the Volume Recovery check box prevents cluster activity churn and reduces the risk of capacity issues due to over-replication during an outage or upgrade.
Important: Do not suspend volume recovery indefinitely, as this would hamper one of Netmail Store's primary data protection layers.
This option allows you to select either a full performance mode or a power-saving mode.
- Full Performance Mode disables the power-saving settings. All nodes remain active (never idle) and volumes become idle after at least six minutes of no I/O activity.
- Power Saving Mode option enables the Sleep After and Wake After settings.
- Sleep After (the
power.sleepAfterparameter) sets the length of time without SCSP activity before health processing is paused and the node displays as Idle in the Admin Console.
- Wake After (the
power.wakeAfterparameter) sets the length of time a node remains idle before the health processor is reactivated.
Note: If your cluster is in constant use (24x7) or if uninterrupted feed restarts are critical for your operations, use Full Performance Mode.
Cluster Tenants Setting
Tenants are secure domains that live in a Netmail Store storage cluster. Similar to a storage facility containing multiple storage units, tenants contain multiple buckets that allow you to store unstructured data objects into specific categories, such as documents, photos, and videos.
For information about managing tenants, see.
Feeds is an object routing mechanism in the Netmail Store storage cluster that uses intermittent channel connections to distribute data to one or more targeted ElasticSearch indexers or target clusters. The source cluster processes all UUIDs and names stored in the source cluster based on your feed configurations. This section lists the feeds that are configured for the source cluster. For information, see.