I do have one test pool which I can not select here.
My pool (since this might be important info), is a single SSD attached through USB interface. I do get warning that my pool is attached through sda device. Is this expected (and can be overriden and if yes, how) or I am missing something (I did restart machine, but nothing changed).
p.s. why are instances marked as âexperimentalâ, is this only related to beta/RC and is expected to be removed with 25.04 release?
Experimental label is because feature still be be hardened and developed more before the 25.10 release this fall, where it becomes more âEnterprise worthyâ. AKA there may be rough edges not yet dealt with, like you just discovered.
This also means that if you already have pools with spaces in their name and wish to use that pool with instances I will need to rename pools (and likely check/recofnigure everything else which is tied to pool name such as shares, snapshost, replications etcâŚ).
Right. Generally I think this should be fairly uncommon situation, but painful if you get stuck in it. I think most admins have an aversion to spaces because the way that scripts can backfire if you mess up escaping strings.
Thanks, I guess I should spend more time examining Incus and its relationship with KVM, as well as understanding what new features/possibilities will be avaliable (I freshly moved from hyper-v, so everythings feels new ). Current KVM portion of GUI feels really rudimentary and I guess Incus will bring more (tuning) options.
Re: Experimental - as I understand, Incus by itself should be quite stable and KVM is proven, so when you talk about rough edges this is likely on how this components interact (truenas with incus with kvm)?
It also feels slightly confusing because docker/incus appear to overlap (somewhat). I have some VMs which could likely benefit from moving to Incus and/or docker (it all depends on networking part which is currently very limited for my liking).
p.s. I really like T3, great content with saturday morning coffee
Coming from windows and with strong OCD urges, I have a real dislike for dashes and underscores, but I get it (and will take this as general guidance in the future).
Yep, its a brand new major feature. Those take a release cycle to fully get fleshed out and bake. Its not so straight-forward as a single on/off checkbox that does a thing. Weâre going to find a lot of unique and weird setups over the next few months Iâm sure.
Similar but different. Lots of use-cases where each option shines better than its cousins Last episode of T3 we went into a lot of that in detail.
On my test system I have re-created my test pool now without spaces and it works correctly, so that portion is maybe not solved but is understood.
Fortunately, VMs I currently use are quite small and it would be fairly easy to move them to another pool, recreate pool without spaces and return them back (I am just a bit depressed that I will have to recreate all snapshot and replication tasks).
p.s. I also noticed a brief sentence in latest T3 podcast/videocast that we will be able to use any pool for Incus (I was a bit confused as why we can choose only one pool), and also there is a question of backup which is not obvious in RC (also mentioned)
Backend API changes merged. UI support forthcoming. Basically, you can define an array of zpools to use for incus storage under the storage_pools key.
When you create an incus volume, instance, etc, the API now has an optional key storage_pool that allows specifying the pool in which to create it.
This allows for configurations where, for instance, VMs are on a different pool than containers. Locating containers in pools where external data is located (for instance media dataset) is interesting from block-cloning / efficiency perspective.