Put default and configured timeouts to InstanceState for using the timeouts while Observe calls

[sergenyalcin]

This PR puts default and configured timeouts to InstanceState for using the timeouts while Observe calls. We had already passed the timeouts to InstanceDiff. This covers the Create, Update, and Delete calls. However, we also need to put the timeouts to InstanceState for' Observe' calls.

We caught the issue while working on the provider-gcp storage.Bucket resource. There is a retry mechanism for Observe calls in the terraform provider for the following reason:

There seems to be some eventual consistency errors in some cases, so we want to check a few times to ensure it exists before moving on.

Also, if we do not set timeouts to resource, the TF provider sets the timeouts to 20m by default. When a user tries to create a new bucket, this operation is stuck for 20m on a retry loop in the Observation state and does not return anything. So, this blocks the creation of this resource.

It seems we have already observed this issue before and set a timeout here for the storage.Bucket resource. To activate this overridden timeout, we must pass these values to the InstanceState, as I mentioned above.

Description of your changes

Fixes #

I have:

• Read and followed Upjet's contribution process.
• Run make reviewable to ensure this PR is ready for review.
• Added backport release-x.y labels to auto-backport this PR if necessary.

How has this code been tested

Tested in provider-gcp by using the storage.Bucket resource manually.

[sergenyalcin]

@ulucinar As we discussed offline, first, I put the timeouts to the Connect function. Then, I checked the situation on the Observe before setting the state. It seems we do not override the timeouts back. I mean, the timeouts remain as we set it. However, being on the safe side, we may consider adding the timeouts again in the Observe function.

[ulucinar]

Thanks @sergenyalcin, lgtm.