Israel Fruchter

`SCYLLA_PRODUCT` is not needed anymore, it can assume it from the version use, or from the branch names. `SCYLLA_ARCH` is expected to be automatically set, base on the system arch,...

Yes, it would need to be applied again, and fixed

> > @aleksbykov still draft? > > @soyacz it is in staging now. working with it @temichus it was in staging month ago ? can someone attend this one ?

we have a latte case that supersede this one, closing for now

> Just need to remove the new addition of "nemesis_add_node_cnt: 3" since it's a default now. > > I would also run the 1tb case, for 12h to make sure...

It does come handy in cases where you have resource issues, or when using spots. For cases no using multiple availability zone, it's actually helpful to have this setting in...

@juliayakovlev this is kind of connected to what you are working on for GCE, i.e. address part of the problem. can you takeover this one ? and test it ?

> @roydahan @soyacz @fruch , a lot of effort went into this PR... and it is not done... > in one hand, to resolve the conflicts, and finish it, may...

@soyacz you think this one is still a good basis for the work needed ? or we'll need to redo it from scratch ?

> I agree. Do we have somewhere the relevant QA-task? https://github.com/scylladb/qa-tasks/issues/414