Hi,

Hi,

I would like you guys to fix the **ibm_pi_instance** resource.

background (via GUI)
1. Lets assume a new pvs service is created.
2. Boot images section is empty
3. Click create instance--->Select HANA or netweaver os image--->Select SLES/RHEL Image.
4. Click create

What happens in GUI background:
When create is triggered, it checks if there is a boot image existing already under boot image section. Since it doesnt exist it will make a copy of boot image and thus a entry is seen in boot image section. Using this boot image an instance is created in chosen tier of choice.
Next time you want to create a new instance with same image:
GUi background checks if the image exists in boot images section and it finds it . it doesnt make a copy , instead uses this image and creates instance.

Power of GUI: Since there is a copy of image from image catalog done automatically , we can deploy instances in any tier

Terraform:
1.using data block catalog_images, we get the id of image.
2. we pass this to ibm_pi_instance and create.

**terraform is not creating a copy of image automatically, it is doing something different and we can create an instance only in same tier as the catalog image tier otherwise you get an error.**
Able to create an instance by specifying cores, mem, etc.
**Not able to create an instance by setting pi_sap_profile_id attribute** and you get the error put in comment below

Power of Terraform: extra copy is avoided.
Cons of terraform: **Cannot create instance in tier of our choice**

Suggestion: **Please change the module ibm_pi_instance working so that copy and check of catalog images is done automatically in the backend.**

_Originally posted by @surajbharadwaj in https://github.com/IBM-Cloud/terraform-provider-ibm/issues/3642_