I can't repro this except in a larger cluster than I can do reasonably in my virtual environment. I will be upgrading to M after the summit. The code looks the same in M however.
Is there a reason why we cleanup expired memcache connections in the same context as token validation and not elsewhere? Seems like it would slow things down.
I can't repro this except in a larger cluster than I can do reasonably in my virtual environment. I will be upgrading to M after the summit. The code looks the same in M however.
Is there a reason why we cleanup expired memcache connections in the same context as token validation and not elsewhere? Seems like it would slow things down.