Remove one or more ranks from an existing Kinetica cluster. All data will be rebalanced to other ranks before the rank(s) is removed unless the rebalance_sharded_data or rebalance_unsharded_data parameters are set to false in the input parameter options, in which case the corresponding sharded data and/or unsharded data (a.k.a. randomly-sharded) will be deleted.
The database must be offline for this operation, see GPUdb.admin_offline()
This endpoint’s processing time depends on the amount of data in the system, thus the API call may time out if run directly. It is recommended to run this endpoint asynchronously via GPUdb.create_job().
Note
This method should be used for on-premise deployments only.
Parameters
Each array value designates one or more ranks to remove from the cluster. Values can be formatted as ‘rankN’ for a specific rank, ‘hostN’ (from the gpudb.conf file) to remove all ranks on that host, or the host IP address (hostN.address from the gpub.conf file) which also removes all ranks on that host. Rank 0 (the head rank) cannot be removed (but can be moved to another host using GPUdb.admin_switchover()). At least one worker rank must be left in the cluster after the operation. The user can provide a single element (which will be automatically promoted to a list internally) or a list.
Optional parameters. Allowed keys are:
rebalance_sharded_data – If true, sharded data will be rebalanced approximately equally across the cluster. Note that for clusters with large amounts of sharded data, this data transfer could be time consuming and result in delayed query responses. Allowed values are:
true
false
The default value is ‘true’.
rebalance_unsharded_data – If true, unsharded data (a.k.a. randomly-sharded) will be rebalanced approximately equally across the cluster. Note that for clusters with large amounts of unsharded data, this data transfer could be time consuming and result in delayed query responses. Allowed values are:
true
false
The default value is ‘true’.
aggressiveness – Influences how much data is moved at a time during rebalance. A higher aggressiveness will complete the rebalance faster. A lower aggressiveness will take longer but allow for better interleaving between the rebalance and other queries. Valid values are constants from 1 (lowest) to 10 (highest). The default value is ‘10’.
The default value is an empty dict ( ).
Returns
A dict with the following entries–
The number assigned to each rank removed from the cluster. This array will be empty if the operation fails.
Additional information.