theking8713

I second this, I don't hate that hexos trips out and alerts me to an unsecure drive state when I add single drive vdevs but since I had to do so in the truenas UI I clearly know enough to be able to do so, so perhaps these warnings are unnecessary as Im likely fully aware of the risks... For my specific application, I really am not interested in redundancy at all, at least not with this operating system, I have several servers deployed in my homelab with close to a petabyte of teared storage solutions were I have plenty of redundant targets and also an offsite hyper compressed archive of the vitals and also cloud targets for configs and documents were it makes a ton of sense to use expensive cloud storage... My needs for hexos is to have an easily deployed on commodity hardware hybrid open target for samba filesharing and app hosting for everyone else that I don't necessarily want to provision privileges to very very expensive resources already in place and I will also use it myself for personal use were I may want to share that space with these other people, like media and games and other files and I can escelate the backup of that shared target to a secure archive in my more enterprise setup... No I want to be able to deploy a low cost mini itx system that only has 4 3.5in HDDs and a single 2.5in sata SSD and 2 m.2 nvme targets were ill have approximately 64TB of space with the 3.5s and a 4tb sata SSD and 2 2tb nvme targets all on a low cost AliExpress special itx motherboard that has 10gb copper via AQC113 Ethernet nic and dual 2.5gb nics via Intel i226 and a basic but sufficient for the application Intel n100 chip were I have enough bandwidth to 2 nvme SSDs to act as a high speed target for que and a semi fast sata SSD for split partitioning for slog and meta and extra cache to supplement the 32gb(the board only has a single DDR5 slot and is limited to 32gb at 4800mhz) of RAM that is primarily leverage as a write cache and also some scratch space for small file muiti user access like document sharing and also space for application folders and any requisite files needed for any VM we may want to tinker with collectively and also the hexos install with be on a small 256gb sata SSD because it is only used for the base install... I'm ok with a little capacity loss by using a raidz2( I think it's called raidz2 for a 3+1 vdev) for some peace of mind that if I lose one of the 16tb drives due to failure I can replace it and rebuild the target in case catastrophe happens between the failure before the scheduled rsync opp but that only for the 3.5in HDDs I don't need any redundancy for any of the solid state targets and I'd like to not be screamed at by the command deck when it detects that there are single device vdevs in the configs and just makes them unaccessible all together so here's hoping this is remediated fairly quickly as again I understand the risks, I don't need any handholding and if I lose data due to my own carelessness that is on my and a risk I was willing to take otherwise I wouldn't have configured it that way deliberately in the harder to use truenas UI... Just sayin

totally open query... looking for thoughts... i do this kind of thing all the time on generic linux installs were i have some overbuilt server and i've obviously put way to many nics into aforementioned servers and workstations and then i have to instantiate a use case for seed unnecessary nics and i don't seem to be keen on wasting nics or there implied bonded bandwidth on redundancy and yada yada on and on and needless to say over the years i have built out an overkill 100GB QSFP local network (not trying to justify my own misbehavior here, BUT this has proven to be remarkably useful in instantiating a pooled stupidly fast kind of "ramdisk" across an entire home that can be accessed as if its local on any workstation equipped with the capacity at these speeds via iscsi... so ya, one can image the use cases for such a pain in the a$$ of a feature iv implemented...lol) and a sfp+ 10gb local network and a sometimes disappointingly unreliable copper 10gb local network that chains up to a 2.5gb level that is a subgroup of a 1GB gateway up to a only 1GB service lol... i'm looking to utilize some of this nonsense of a spiders nest of fiber and copper iv unjustifiably ensnared myself in... 😛 ya i'm definitely in over my head considering i've never successfully deployed a base truenas scale install, just always used generic linux distros lol... I fear I'm encouraging feature creep here... plz tell me if this is the case or is some kind of "soft" switching fairly common in off the shelf NAS solutions... idk it's been so long since i've needed to mess with them... i think the last one i had was a 4TB WD my cloud that literally only had a 100MB nic and that was top of the line at that time lol... totally off topic but kind of related... i'm not as much excited about the truenas base except for ACLs... ya linus from LMG mentioned that as well in there coverage of this future product and coming from a "using generic GUI linux distros for server purposes" kind of guy this is EXACTLY the thing i am most in need of a solution for lol... and maybe unity on the systemd vs openrc front... idc which but some clarity is all i ask lol i think at the end of the day, for me and i suspect a lot of people many if not most of these "nice to have" features will be satisfied by us going into the truenas interface and implementing them ourselves and this will simply be a very compelling "foot in the door" and for that even tho i have yet to even test this beta on my own i'm still stoked at the prospect... lol o and i know i cant spell or punctuate to save my life, i'm in my mid thirties and i haven't remediated this known quality yet so plz refrain from roasting me on this... keep the critiques relative to the query,,,thanks

oh... interesting, i totaly forgot i have this itx board from radxa that is a super overkill SBC lolz, hmm... ARM HexOS nas? now that might be the usecase iv been wating for on this board... BUT i do want to preface that by saying its not a priority at least not for me considering how long weve waited on that radxa product just to get a funcional kernel no thanks to the original developers but instead to due to community devs like Heiko Stuebner and thats just to get a few pcie fetures working on a mainline debian install let alone one that would work for a brand new unknown nas os based on a niche enterprise distrobution lol... no lets afford them some leaway and have them put maximum effort into assuring x86 works as intended first... ya?

Ipxe?... Idk if I'm asking over my head here perhaps PXe is something else I genuinely don't know but ipxe boot perhaps via a tftp service is something that can easily be done from truenas If I remember right but also it can easily be done from a router/firewall because that is the DHCP provider but again I might be asking a dumb ? I'm on with that lol