OPE Storage

TIBCO OPE Storage (OPES) needs to store customer model XML and processed cached objects in permanent storage so as to persist data across OPE micro service container restarts.

OPE can store data across any of the following backend storage systems using MinIO object storage
  • File System
  • NFS Storage
  • SAN Storage
  • Amazon S3
  • Azure Blob Storage
  • Google Cloud Storage
The following model types will have 3 buckets each in MinIO for model processing:
  • Product
  • Price
  • Discount
  • Category
  • Rule
Each Model will have the following 3 types of buckets:
  • Input
  • Failure
  • Success
For example, for the product model, the MinIO buckets will be:
  • product-models-failure
  • product-models-input
  • product-models-success

For 5 model types with 3 different types of buckets, there will be 15 buckets for processing model XMLs.

Similarly to store processed cache information, separate cache buckets are created for each model type; therefore, the bucket names for the processed cache will be:
  • cache-products
  • cache-prices
  • cache-discounts
  • cache-categories
  • cache-rules
Note: All bucket names have lower-case characters.
Here is an example screenshot of how the buckets look in the MinIO UI. The MinIO UI can be accessed on http(s)://<host>:<port>:9000.

Note: All the buckets are needed to be created through the MinIO UI to ensure that there are no access permission issues on the buckets/folders on the file system.

The following is an example UI screen with an option (highlighted in the RED box) to create a new bucket.