Closed Xarthisius closed 1 year ago
Base: 92.80% // Head: 92.83% // Increases project coverage by +0.02%
:tada:
Coverage data is based on head (
0cd50e1
) compared to base (fe022c0
). Patch coverage: 95.83% of modified lines in pull request are covered.
:umbrella: View full report at Codecov.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.
This works as expected, but raises questions for me about how these specific events tie to higher-level metrics we can use in reporting. Some of these tend towards technical (instance deletion) and may lack certain details (do we track instances of specific environments if the environment changes?)
It made me think about about what reporting could look like (ala csdms) and led me to an oldie but goodie. Obviously many of these are just reports on existing data, but it might be helpful to think through any additional data we need to collect?
tbh I wasn't even thinking about reporting for funding agencies. I wanted to see answers to questions that I personally have. DELETE /instance
metric in particular keeps the state of the instance being deleted. That's useful to get a % of failures. Currently, I only track taleId
in instance related events, but could just easily add imageId
being used.
"Thinking through" is precisely the point of this PR. Additional suggestions are highly encouraged :)
Added info about container that was running to instance.remove
metrics:
{'_id': ObjectId('639b44cc7c7bb3fb9e8a4664'),
'type': 'instance.remove',
'details': {'id': ObjectId('639b449b7c7bb3fb9e8a4635'),
'taleId': ObjectId('639b41b4e81d22b4bb8402d4'),
'status': 1,
'containerInfo': {'mountPoint': '/var/lib/docker/volumes/639b41b4e81d22b4bb8402d4_kowalikk_5Tb5Yb/_data',
'digest': 'registry.local.wholetale.org/tale/b3fe3003ade99e0d3b07702430c16712:20f366cbe8bed898977c28417230ebcb@sha256:bdf23063af6b67b3590a9ed604aee1ddd7aca08ace2908ca95461c040f2b6e63',
'nodeId': 'i4q72m3jebgjdd076x0ifm7b8',
'volumeName': '639b41b4e81d22b4bb8402d4_kowalikk_5Tb5Yb',
'imageId': ObjectId('639b40ecd8d2b79b6089fd4c'),
'name': 'tmp-nysubzrklvpq'}},
'ip': '10.255.0.2',
'userId': ObjectId('639b41ade81d22b4bb8402d2'),
'when': datetime.datetime(2022, 12, 15, 16, 1, 16, 146000)
}
Arguably this should happen 5yr ago, but here we are... This PR introduces a custom logger that saves records of particular actions in the database. Currently it tracks:
wouldn't it be nice to have those metrics for NSF reporting?
How to test