Closed mkeeter closed 2 months ago
built a tuf repo with this + propolis main + omicron main. threw the tuf repo on madrid. my instance has been "starting" for a suspiciously long time now, so maybe current propolis main + current omicron main or current propolis main + crucible main isn't a good combo.
but before that point, uploading a boot iso worked just fine so i think that means the crucible + omicron + illumos side of things is happy (because it had to get through uploading blocks)
built a tuf repo with this + propolis main + omicron main. threw the tuf repo on madrid. my instance has been "starting" for a suspiciously long time now, so maybe current propolis main + current omicron main or current propolis main + crucible main isn't a good combo.
but before that point, uploading a boot iso worked just fine so i think that means the crucible + omicron + illumos side of things is happy (because it had to get through uploading blocks)
Did you ever track down where this was getting stuck? Do we think it's something to do with the changes here?
built a tuf repo with this + propolis main + omicron main. threw the tuf repo on madrid. my instance has been "starting" for a suspiciously long time now, so maybe current propolis main + current omicron main or current propolis main + crucible main isn't a good combo. but before that point, uploading a boot iso worked just fine so i think that means the crucible + omicron + illumos side of things is happy (because it had to get through uploading blocks)
Did you ever track down where this was getting stuck? Do we think it's something to do with the changes here?
Haven't. I have not had a chance to do another racklette run. I was thinking to try this branch rebased on a crucible commit that's actually been landed into omicron along with a matching propolis to see if that works fine
VM launching work, and a full iodriver perf test completed ok. So I think we are in the clear.
Opening this as a draft right now, to check whether it works on Buildomat's new helios image.