#pyramid

x58 any downsides, other than being for humans?

I keep getting timeout errors with pymongo, and my performance measure of my pyramid shows 93% of time is spent in acquire _thread.lock

What's the stack look like at the time if calling _thread.lock

Jonno_FTW: Also, do you have a lot of views? Pyramid attempts to cache the view lookup, but that should only be done once for each (request iface, context iface and view name) tuple.

And should thus not grow without bounds.

seems unlikely

but yeah gonna need to know what's trying to lock

view caching is the only lock we have in Pyramid. Otherwise it's outside of our control.

The view cache is naughty and updates the registry.

Looking at the code, it shouldn't need a lock though, the lock is not protecting anything.

Other than potentially concurrently updating the same dict... but that isn't an issue.

Since Python interpreter will take care of that.

cpython yes

looks like I had a ridiculously long query

took 12s

it's examining every document in the collection, all 1.2M of them