-
Notifications
You must be signed in to change notification settings - Fork 296
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
Dragonfly preheat not speeding up subsequent image load; errors reported #3674
Comments
@amholler Please provide the full log in |
Hi, @gaius-qi have uploaded a tar that includes the dfdaemon.log files from each of the 3 client instances. Thanks! |
@amholler This log is the time when you downloaded a piece from the remote:
This log is the time when you downloaded a piece from the source:
The faster P2P download speed needs to meet the following:
|
Hi, @gaius-qi Thanks for your analysis! But I'm missing something here: I had issued a successful preheat to all peers, so why does any peer to peer operation need to happen? I thought all peers should have the full image after the preheat from source completed, so image load after the preheat should come directly from the local cache, and hence be very fast. |
@amholler How do you preheat the image to all peers? |
Hi, @gaius-qi , Thanks for responding.
As I mentioned in the first comment in this ticket, I did it as follows:
|
@amholler Please provide Manager, Scheduler and one of the peer logs. |
Bug report:
Expected behavior:
Expect image pull time to be greatly reduced after the preheat. Don't expect the client logs to be filed with errors.
How to reproduce it:
See details in section above.
Environment:
uname -a
): Linux gke-anne-regional-default-pool-23f7047d-0c4w 6.1.100+ [WIP] Implement df daemon #1 SMP PREEMPT_DYNAMIC Sat Aug 24 16:19:44 UTC 2024 x86_64 GNU/LinuxThe text was updated successfully, but these errors were encountered: