we're good to close this ticket.

I can confirm that our slaves are in good use at the moment. We have a total of 150 online slaves, 16 of which are PPC ones and the current slave utilization is at 87% which I would call very good:
http://graphite.phx.ovirt.org/dashboard/db/jenkins-monitoring?panelId=16&fullscreen
Unfortunately some reporting in Graphana is broken, that's why it's showing "1 running build" on the appropriate graph - please ignore that.

The build queue went down from 1500 to 1200 in 1 hour, so I would estimate the whole queue will be processed within 4-5 hours.

Looking at teh oVirt admin UI, hypervisors have around 87% CPU load as well, so adding more slaves will just slow down the others: we're running near hardware capacity at this point.
could you please confirm if there is anything I can do regarding this or we're good to wait for all the builds to complete?

looks like my estimates were off by a factor. we have over 112 patches running in CI ATM: https://gerrit.ovirt.org/#/q/topic:mockito

We also have more slaves, not sure about the exact number but looks like about 100.
how many x86_64 slaves do we have ATM?

Load seem to be due to ~20 engine patches sent at once, each running multiple upgrade jobs... probably not neccessary so maybe we should just abort the upgrade jobs if important tasks are starved atm.

This is a false alarm, we have more then 50 slaves working.

The unused ones are the PPC slaves.