I'll stay with option A, as is part of Google's traditional testing (Canary Test).. Look fot this topic.. more info on their SRE ebook at: https://landing.google.com/sre/book/chapters/testing-reliability.html

I think the answer is B.
A is weird because it is saying deploy changes to a small subset of users BEFORE rolling to production. That is less users testing the changes/

For me question is not really clear :/
A & B could be good reply.
As they speak about performance bugs so maybe there is not enough stress on the staging en => I would say B
If you deploy with canary update, you will get the issue in all case but directly in production so for me not the best option.

BUT GCP is pushing a lot canary update so it could be the good reply but I don't know :/

In Canary Deployment you deploy to a small fraction of your users to check everything’s fine before rolling out to more users.

Increasing the load on the test is not going to find the issue, I will go with the option A.

I think Robert is right
Option B is about increasing the load. The performance bugs may be caused due to the scale and not the load. Question is not pointing out to any such spike or sudden load.

Hi mamh, the fact is "did not see in staging".
Increasing the load is not an option here since there is no problem at all. And in the future it could happen again.
So here we should use canary testing.

why the ans is "A", not "B"