Skip to content

Latest commit

 

History

History
116 lines (80 loc) · 5.07 KB

README.md

File metadata and controls

116 lines (80 loc) · 5.07 KB
Raw

DB2 Warehouse

1. How is DB2 Warehouse used in Maximo Application Suite

DB2 Warehouse forms the datasource for Monitor to store all the application metadata and analytics data provided by the Monitor KPI.

1.1 IoT Data Store Connector and DB2

The Data store connector (DSC) component in IoT is used to store device event and state data in DB2 warehouse. Connecting IoT to a DB2 warehouse service allows you to store and access the device/state data. When the Monitor application is deployed, the configured DB2 service is automatically connected to IoT as part of Monitor deployment procedure.

In IoT, the destination resource is used to represent tables in DB2. Creating/Deleting a destination in the platform will attempt to create/delete the table in the target DB2 service.

The configuration property on the destination resource will contain metadata that describes the structure of the DB2 table. On a create of a destination resource, this metadata will be used to generate the CREATE TABLE SQL statement that will be executed against the target DB2 service. For example:

{
  "name": "TEMPSENSOR_LI",
  "type": "db2",
  "configuration": {
    "columns": [
      { "name": "temp", "type": "float", "nullable": true }, 
      { "name": "hum", "type": "float", "nullable": true }
    ]
  }
}

1.2 Data Structure in DB2

Monitor creates a dedicated schema for every workspace Id, schema is named as <workspaceId>_MAM. All the tables created for a workspace is created under its own schema.

Activated Logical Interface

The Monitor application is notified whenever a user configures device state model (schemas, event types, physical and logical interfaces) or activates a logical interface within IoT, on receiving this notification the Monitor application will create new tables and metadata configuration within DB2 Warehouse, and automatically configure forwarding rules in IoT to direct data flow into the table.

As a result, a table with name IOT_<logical_interface_name> is created in DB2. The properties of the activated logical interface is used as the metadata along with default information of the LI like device type ID, device ID and timestamps in UTC.

For example: In workspace test, for an activated logical interface dt0_LI's schema as shown below for device type dt0 and device d0,

{
    "$schema": "http://json-schema.org/draft-04/schema#",
    "type": "object",
    "title": "schema-title-43702536",
    "description": "",
    "additionalProperties": false,
    "properties": {
        "temp": {
            "type": "number"
        }
    }
}
  • Monitor creates a table with name IOT_DT0_LI in schema TEST_MAM.
  • The Table's columns are based on the property of the LI schema, so in this case, TEMP with Integer type and other default properties like DEVICEID, DEVICETYPE, RCV_TIMESTAMP_UTC and UPDATED_UTC columns are created.
  • Monitor also creates the forwarding rule to direct the events into this destination.
  • So any state update received for this LI will be forwarded to the DB2 by the DSC component in the platform.
  • When the device publishes temp reading 15 and 16, we can see two rows with those values in the below image.

DB2 view for the logical interface dt0_LI

Device Metadata

Similar to the above scenario, when a device's metadata/info is created or updated, Monitor is notified and a table IOT_<logical_interface_name>_CTG is created in the workspace's schema(<workspaceId>_MAM). Each metadata property will be a seperate column in the table.

2. Supported Versions

The following DB2 Warehouse version releases are supported:

  • 11.5.X

3. Resource Requirements

Use the online tool to calculate the cluster resource requirements

4. Installation

??? abstract "Prerequisite Actions" - Install CloudPak for Data

  • oc login to your OpenShift cluster
  • Go to the directory you placed the repo.yaml
  • Run the cpd adm command: ./cpd-linux adm -r repo.yaml -a db2wh -n zen --apply --verbose --accept-all-licenses
  • Run the cpd install command:
./cpd-linux -s repo.yaml \
  -a db2wh \
  -n zen \
  -c <storage_class> \
  --verbose \
  --target-registry-username openshift \
  --target-registry-password $(oc whoami -t) \
  --insecure-skip-tls-verify \
  --cluster-pull-prefix image-registry.openshift-image-registry.svc:5000/zen \
  --transfer-image-to image-registry.openshift-image-registry.svc/zen

5. Supported Storage

System and backup storage, and user storage both require a storage class that supports the ReadWriteOnce access policy.

OpenShift Container Storage 4.3

  • System and backup storage: ocs-storagecluster-cephfs (select 4K sector size)
  • User storage: ocs-storagecluster-ceph-rbd (select 4K sector size and ReadWriteOnce)

IBM Cloud

IBM Cloud File Storage (ibmc-file-gold-gid storage class) & Portworx:

  • System and backup storage: portworx-db2-rwx-sc (select 4K sector size)
  • User storage: portworx-db2-rwo-sc (select 4K sector size and ReadWriteOnce)