-
Notifications
You must be signed in to change notification settings - Fork 693
Zerg dance: writefifo race condition #1486
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
Comments
FWIW -- I have wrapped the gist in a repo so that you can do this:
Hope it helps someone ... |
rdeioris
pushed a commit
that referenced
this issue
Mar 16, 2017
The spinningfifo action has been added to the core. @xrmx safe for backport |
unbit
added a commit
that referenced
this issue
Mar 25, 2017
Modify writefifo to spinningfifo solved this problem. Thanks @pennersr |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I have tried to setup a Zerg dance by following the "Art of Graceful Reloading", and could not get it working reliably out of the box when the number of processes is >1.
The problem is that the documentation instructs you to send "1" to the "new.fifo", but when the master is still booting up the workers the "new.fifo" may not be present yet. Example output:
A proper run would like like this (here, the workers have all been booted an the write to new.fifo succeeds):
Apparently, more people have been running into this:
The workaround based on the spinningfifo works for me:
Here...
... it was stated that:
It would be nice if this were tackled -- it would have saved me a lot of time.
The text was updated successfully, but these errors were encountered: