search for: jasonsydes

We ran into some problems accessing a freshly created worker, and had to insert a sleep of 1 second to get it to work. It looks like Ezra and skaar already know this: from middleman.rb: # HACK: there is a race in the worker/slave creation, we # currently need to sleep between create. sleep 0.1 You might want to increase it to 1 second for those of us on slowly mac books.

What are people are doing when they need to deploy a new code base? You can''t just restart backgroundrb, because then you lose any long-running backgroundrb processes. But then you''d conceivably have to wait several hours for all your procs to complete before pushing out the new code base. I was originally thinking that during code deploys, we''d start up a second