Profile + Optimize performance of saving volume buckets / updating segment index

[fm3]

When the server is busy, we see some volume update requests becoming very slow (>10m). Can we do something to improve performance?

[fm3]

I’d consider this fixed by #8460 #8482 and #8489 (also vaguely related is the upcoming #8469 )

Of course we can always optimize further, let’s do so if we notice more slow requests.