Preparing to install Watson Query

Plan and prepare to install the Watson Query service.

Watson Query pods

The Watson Query service instance has two main pods:

Head pods

The head pod is c-db2u-dv-db2u-0.

The Watson Query head pod runs the Watson Query head component (also known as the engine).
Worker pods

Worker pods are c-db2u-dv-db2u-x where x is 1 or greater.

In Watson Query, the term worker pod refers to a pod that runs one Watson Query worker component. You can allocate multiple worker components to the Watson Query service instance. These components are effectively multiple c-db2u-dv-db2u-x pods.

You must not confuse Watson Query worker pods with compute nodes, which are the physical nodes that make up the Red Hat® OpenShift® cluster. For more information about cluster compute nodes, see Architecture for Cloud Pak for Data.

System requirements

Ensure that you meet the service requirements that are listed in System requirements. The Watson Query service runs on x86_64 hardware only.

The Watson Query service is provisioned to any compute node in the Cloud Pak for Data cluster that has the specified resources (cores, memory, and ephemeral storage) available.

Additionally, the Cloud Pak for Data cluster must accommodate the initial provisioning request for Watson Query service pods.

Important: If you try to provision a Watson Query service instance and you don't have enough resources, the provisioning fails.

Scaling

You can scale the Watson Query service up and down at any time after you provision it. For more information, see Scaling Watson Query.

Work with IBM Sales to get a more accurate sizing based on your expected workload.

IBM Sales helps you estimate the total demand for Watson Query. The service then redistributes resources internally. When you provision Watson Query.

Storage requirements

Ensure that you meet the storage requirements that are listed in Storage requirements.

At a minimum, the persistent storage must meet the following requirements for Watson Query.
  • Persistent volume for Watson Query engine node is 50Gi.
  • Persistent volume for Watson Query caching is 100Gi.
  • Persistent volume is XFS formatted.

If you are using an NFS storage class, you must ensure that the NFS export is set to no_root_squash before you install. For more information, see Setting up NFS storage. To troubleshoot issues, see SQL6031N error in the db2nodes.cfg file in Watson Query.

External libraries

External libraries (that is, libraries that are not included in the Watson Query service) are stored on a persistent volume. Watson Query automatically creates persistent volume claims during the provisioning process.

This storage is the same as the persistent volume for the Watson Query head pod.

The persistent volume claim for external libraries must have at least 50 GB available.

Cache storage
A data cache holds temporary data that is used frequently. By using a data cache, you can reduce processing and loading time when you use this data.

Kernel parameter settings

To ensure that Watson Query can run correctly, you must verify the kernel parameters. For more information, see Changing kernel parameter settings.

CRI-O container settings

Complete the steps in Changing CRI-O container settings to set the pids_limit parameter.

Ensure that the pids_limit parameter is equal to or greater than the minimum value that is a prerequisite for Cloud Pak for Data.