-
Notifications
You must be signed in to change notification settings - Fork 37
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
send cluster status to den when running runhouse status cli command #1097
Conversation
This stack of pull requests is managed by Graphite. Learn more about stacking. |
I'd rather not have this logic duplicated between the ClusterServlet and main.py, and generally we try to have interfaces with ClusterServlet and EnvServlet go through the obj_store, not directly from main.py or HTTP_server. Can we just modify the |
0990747
to
76921a1
Compare
76921a1
to
73a3bb6
Compare
73a3bb6
to
1688d7b
Compare
1688d7b
to
e0ff18c
Compare
e0ff18c
to
0509b93
Compare
b56c01a
to
4569713
Compare
4354a26
to
de74386
Compare
@@ -201,6 +202,37 @@ def test_fn_to_docker_container(self, ondemand_aws_cluster): | |||
#################################################################################################### | |||
# Status tests | |||
#################################################################################################### | |||
|
|||
@pytest.mark.level("minimal") |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
these tests should prob be in test_cluster
no? doesn't really need to be limited to ondemand
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
yeah, but we are using cluster.teardown()
which is an on-demand cluster method. Runhouse Cluster does not have such method. Moreover, seems like runhouse stop
is not a part of runhouse cli API (not in the docs - is it on purpose?), so not sure if we should use it in the test to this use-case.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
why do we need to do teardown? i see those tests are being skipped anyways maybe we can remove them
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
to test that once we call cluster.teardown() the updated status is sent to den. We could delete it for now since they are skipped, but I think it better to add them to a separate test suite that runs tests related to cluster termination (I think we've previously talked about it). wdyt?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
you mean confirming that when you call teardown the status gets updated in Den? maybe i'm missing something but i don't see that in this test?
b4e8562
to
6e1af83
Compare
) | ||
logs_end_line = cluster_status.get("cluster_config").get("end_log_line", None) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
what about comparing the start and end lines before and after the status call? wouldn't that be a way to confirm the status update worked without actually having to test the call to Den?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
we could do that, but it this case we'll have to wait 1 min to ensure that the cluster check loop were executed during the test. This will cause quite latency in executing the tests....
In this case i think that calling den is less "harmful". wdyt? @jlewitt1
6e1af83
to
f906fb9
Compare
f906fb9
to
a1e13a8
Compare
a1e13a8
to
da456c2
Compare
da456c2
to
4fb129a
Compare
No description provided.