fix(discovery): when endpoint_slices updated, old endpoints all flush_all/flush_expired

[dongjiang1989]

Description

Fixes # (endpoint muti slices)

Bad case：
If the number of nodes of a certain service is too large, for example, it is split into two endpoint slices:
slice1: There are pod1 and pod2.
slice2: There are pod3 and pod4.
If the pod is re-scheduled and only slice1 or slice2 is updated, the upstream will only be updated to pod1 + pod2 or pod3 + pod4.

When watch_endpoint_slices_schema is true(endpoint_slice model), remove informer.pre_list and informer.post_list

Checklist

• I have explained the need for this PR and the problem it solves
• I have explained the changes or the new features added to this PR
• I have added tests corresponding to this change
• I have updated the documentation to reflect this change
• I have verified that this change is backward compatible (If not, please discuss on the APISIX mailing list first)

[shreemaan-abhishek]

please add test cases and explain the changes in detail and how it solves the issue.

[dongjiang1989]

please add test cases and explain the changes in detail and how it solves the issue.

Thanks @shreemaan-abhishek .
Add the changes done.

[dongjiang1989]

/ok-to-test

[shreemaan-abhishek]

please add test cases

[dongjiang1989]

/ok-to-test

[dongjiang1989]

please add test cases

Thanks @shreemaan-abhishek .
Add e2e cases done.