-
Notifications
You must be signed in to change notification settings - Fork 79
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
Rockset migration - move upload for queue_times_historical from rockset to s3 #5398
Conversation
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
@@ -48,6 +48,7 @@ SELECT | |||
COUNT(*) AS count, | |||
MAX(queue_s) AS avg_queue_s, | |||
machine_type, | |||
CURRENT_TIMESTAMP() AS _event_time |
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.
Maybe a comment on why we need to use CURRENT_TIMESTAMP as _event_time here. I guess that this is because Rockset uses the time when it ingests the data as _event_time. So, re-importing records from S3 will set them all to the same _event_time?
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.
Yes _event_time is the time records get added to rockset if it is empty https://docs.rockset.com/documentation/docs/special-fields#the-_event_time-field
I decided to make a different field entirely so theres less confusion between the two
…low (#5411) Role defined in pytorch-labs/pytorch-gha-infra#441 Related * #5398
…ield instead of _event_time and fix env variables in update queue times workflow (#5412) Related to #5398 Change historical queue times query for HUD metrics page Looks good <img width="847" alt="image" src="https://github.com/pytorch/test-infra/assets/44682903/3376331c-01f6-45e9-a22c-da359ce6df89"> Also change the workflow so we can use workflow dispatch and so that it has the env vars
Use the machine type and time as the key