[core] Counting number of status updates in transition_deploy

[justonedev1]

@knopers8 Do you think, that this can work for making environment deployment fail faster? I am not sure whether number of leafs gathered from LeafWalk can be comparable to the total amount status updates that we are receiving.

I tested that only on my openstack with 5 tasks. But if you think that it is worth investigating, I can take a look at staging

[knopers8]

I'm afraid we are relying here on a quite far-fetched assumption that there is always just one status update per leaf. Even if this is correct now, we risk breaking this optimization if something changes about it in the future.

I would rather investigate if we could have a stupid channel listening for any deployment failure of a critical task and abort accordingly this loop in case that happens.