Skip to content
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

Speed up launcher startup #1567

Merged

Conversation

RebeccaMahany
Copy link
Contributor

  • Once() can take up to 5 seconds to run -- don't block on running it
  • We can't use the Secure Enclave right now -- stop waiting for up to a second to access it

ee/agent/keys.go Outdated Show resolved Hide resolved
zackattack01
zackattack01 previously approved these changes Jan 30, 2024
James-Pickett
James-Pickett previously approved these changes Jan 30, 2024
Copy link
Contributor

@James-Pickett James-Pickett left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nice

checkpointer := checkups.NewCheckupLogger(logger, k)
checkpointer.Once(ctx)
go checkpointer.Once(ctx)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The prior way was intentional -- it was intending to get some of the identifying information into the logs before any early crashes might occur. It's fine to revisit, maybe we don't need it with all the things we've subsequently added.

Though I think this begs the question -- why does it take 5s to log, and should we add go routines downstack?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I figured, good to get identifying info into the logs as soon as possible, but no need to block on it. And I think I remember we added this around when we were still working through the Monterey bug issues, so thought this might not be as critical anymore -- but I could be misremembering.

I will check to see which of the checkups is causing the slowness!

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I wonder if we should just async that one check. Not sure

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

That seemed like maybe overly complicated to coordinate an exclusion for that one checkup for the checkpoint only, but not doctor/flare

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thinking about it... Maybe this simple thing is good enough for now.

We probably do want to keep this in the log checkpoint, it provides an interesting point-in-time view.

I suspect that the theoretical correct state would be to make the timer log checkpoints use a different set of debuggers than the initial startup ones. And I'm not sure it's worth doing that this moment

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Agreed!

ee/agent/keys.go Show resolved Hide resolved
@RebeccaMahany RebeccaMahany added this pull request to the merge queue Jan 31, 2024
Merged via the queue into kolide:main with commit 7e626b8 Jan 31, 2024
26 checks passed
@RebeccaMahany RebeccaMahany deleted the becca/launcher-startup-speedup branch January 31, 2024 18:51
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants