Production running *really* slow
O-I opened this issue · comments
I have no idea what's going on. There are no errors in the Heroku logs, I'm running on a fast network, and performance is dead slow regardless of whether I use Chrome or Safari. There is no issue in development. Investigating...
Looks like it is related to this: mislav/will_paginate#400
Just when I thought this was finally resolved. [sigh]
Now I'm completely confused. Using my own patch that mimics nazgum's fork resulted in production running as usual. I expected switching to edariedl's patch to be an insufficient fix for my purposes (I tried the same basic fix last night — it did not solve my performance woes). Surprisingly, it seems to work fine. So whence did the performance issue come?
It''s possible it might have something to do with the rake task I edited. Maybe I'll wait until Monday morning after all the rake tasks have been run at least once by the Heroku scheduler to see if the sluggish performance reappears.
It's also possible this was just a freak occurrence on Heroku's end. There were no discernible errors in any of the logs — performance was just enormously slow and seemed to back up on requests in my impatience.
Well, tomorrow is another day.
More evidence that this was just a Heroku performance issue: Increased HTTP Latency
OK. Seems like everything is back to normal. Guess it was just a Heroku performance blip.