Enforce unique contraints on object names/labels in the database.

[xuhang57]

Fix: #946

Potentially, we should think about whether there are any other constraints.

And this might break lots of tests. Consider this is an on-going implementation

[xuhang57]

Some help would be much appreciated.

I am suspecting that because this is called many time:

https://github.com/CCI-MOC/hil/blob/45f218d272a4da8cd21694bcf1de0783a165b756/tests/unit/api/auth.py#L612

so the node name is not unique in the CI database?

[coveralls]

Pull Request Test Coverage Report for Build 1693

• 1 of 1 (100.0%) changed or added relevant line in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage remained the same at 59.338%

---

Totals
Change from base Build 1686:	0.0%
Covered Lines:	2097
Relevant Lines:	3534

---

💛 - Coveralls

[coveralls]

Pull Request Test Coverage Report for Build 1729

• 11 of 19 (57.89%) changed or added relevant lines in 3 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage decreased (-0.07%) to 59.273%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
hil/migrations/versions/264ddaebdfcc_make_labels_unique.py	6	14	42.86%
<!--	Total:	11	19	57.89%	-->

Totals
Change from base Build 1686:	-0.07%
Covered Lines:	2103
Relevant Lines:	3548

---

💛 - Coveralls

[zenhack]

Could you also write a migration script? Other than that, LGTM.

[naved001]

project names, switch names, network names are unique too.

[naved001]

Honestly, I'm not sure it's worth the effort to properly support database downgrades, and we've only been adding the function at all because alembic expects it.

I agree.

[zenhack]

I'm happy. If everyone's in agreement re: coveralls we can just merge.

[naved001]

Cool, merging then.