Evaluate usage of materialized views for the scoreboard about enoengine HOT 7 OPEN

Is this really necessary?
It feels hard to understand, tweak, and maintain.
The bus factor is already low for the engine, but this reduces it to < 1.

from enoengine.

That depends on how long the round time calculation is allowed to take for a large number of teams and submissions, obviously 😛

Do you think faust's code looks so bad?

from enoengine.

I heard not everybody at Faust knows what's going on there.
The round time calculation should not be longer than the round time...
What's the current perf?

from enoengine.

What's the current perf?

We can answer that question only for a given number of teams, services, submitted flags and hardware 😇

Let's hope we can do a stresstest on the next hetzner test with a few hundred teams, I am really curious 🍿

from enoengine.

The calculations could be done doing some mocking, right? Didn't Moritz write something last year?

from enoengine.

Yeah, but I still need a system with a decent ssd and some memory for a fair test

from enoengine.

The OpenNebula host has a Xeon E5-2609 v4, 32 GB RAM but unfortunately only an HDD. But since MySQL should cache most of the stuff in RAM this should still be somewhat usable, right?

from enoengine.