Query weak consistency policy

The query weak consistency policy is a cluster-level policy object that configures the weak consistency service.

Management commands

• Use .show cluster policy query_weak_consistency to show the current query weak consistency policy of the cluster.
• Use .alter cluster policy query_weak_consistency to change the current query weak consistency policy of the cluster.

The policy object

The query weak consistency policy includes the following properties:

Property	Description	Values	Default
PercentageOfNodes	The percentage of nodes in the cluster that execute the query weak consistency service (the selected nodes will execute the weakly consistent queries).	An integer between 1 to 100, or -1 for default value (which is currently 20%).	-1
MinimumNumberOfNodes	Minimum number of nodes that execute the query weak consistency service (will determine the number of nodes in case PercentageOfNodes*#NodesInCluster is smaller).	A positive integer, or -1 for default value (which is currently 2). Smaller or equal to MaximumNumberOfNodes.	-1
MaximumNumberOfNodes	Maximum number of nodes that execute the query weak consistency service (will determine the number of nodes in case PercentageOfNodes*#NodesInCluster is greater).	A positive integer, or -1 for default value (which is currently 30). Greater or equal to MinimumNumberOfNodes.	-1
SuperSlackerNumberOfNodesThreshold	If the total number of nodes in the cluster exceeds this number, nodes that execute the weak consistency service will become 'super slacker', meaning they won't have data on them (in order to reduce load). See Warning below.	A positive integer that is greater than or equal to 4, or -1 for default value (currently no threshold - weak consistency nodes won't become 'super slacker').	-1
EnableMetadataPrefetch	When set to true, database metadata will be pre-loaded when the cluster comes up, and reloaded every few minutes, on all weak consistency nodes. When set to false, database metadata load will be triggered by queries (on demand), so some queries might be delayed (until the database metadata is pulled from storage). Database metadata must be reloaded from storage to query the database, when its age is greater than MaximumLagAllowedInMinutes. See Warning and Important below.	true or false	false
MaximumLagAllowedInMinutes	The maximum duration (in minutes) that weakly consistent metadata is allowed to lag behind. If metadata is older than this value, the most up-to-date metadata will be pulled from storage (when the database is queried, or periodically if EnableMetadataPrefech is enabled). See Warning below.	An integer between 1 to 60, or -1 for default value (currently 5 minutes).	-1
RefreshPeriodInSeconds	The refresh period (in seconds) to update a database metadata on each weak consistency node. See Warning below.	An integer between 30 to 1800, or -1 for default value (currently 120 seconds).	-1

Important

Prefetch operation requires pulling all databases metadata from Azure storage every few minutes (in all weak consistency nodes). This operation puts a load on the underlying storage resources and has impact on cluster performance.

Warning

Consult with the support team before altering this property.

Default policy

The default policy is:

{
  "PercentageOfNodes": -1,
  "MinimumNumberOfNodes": -1,
  "MaximumNumberOfNodes": -1,
  "SuperSlackerNumberOfNodesThreshold": -1,
  "EnableMetadataPrefetch": false,
  "MaximumLagAllowedInMinutes": -1,
  "RefreshPeriodInSeconds": -1
}