-
Posts
1582 -
Joined
-
Last visited
-
Days Won
88
Content Type
Profiles
Forums
Articles
Blogs
Store
Everything posted by Mobius
-
Questions that googling/scanning the forums aren't helping.
Mobius replied to kdcampo's topic in Roadmap & Feature Requests
yeah it might be time for new hardware. I would also suggest not keeping your only copy of important data on this machine if you intend to keep using it as is. xeons are excessive if the only reason you are picking them is for less of a headache. for most users a used cpu and mobo is good enough. heck in your case i might even look for a replacement motherboard on the cheap. -
Expanding Storage Pools in Hexos – Support for Drives of Different Sizes?
Mobius replied to bonecrusher's question in OS & Features
since you are just replacing 2 drives, have you considered just plugging your 2 2tb drives and 2 8tb drives at the same time, copying over your data then removing your 2tb drives? -
im no expert on permissions (and i don't even use HA) but, when you remade the data sets did you select the apps preset?
- 7 replies
-
- hexos
- installation
-
(and 2 more)
Tagged with:
-
Expanding Storage Pools in Hexos – Support for Drives of Different Sizes?
Mobius replied to bonecrusher's question in OS & Features
at this time it will have to be done via the underlying truenas interface -
i would say if you don't intend to do ai workloads, skip the gpu and get one of the few xeons with an igpu. just by existing a dgpu like your 1660 super will be using like 10-20w at idle. intel quick sync transcoding is actually really good and nothing to scoff at. I think even now it is considered the gold standard for low power transcoding
-
Unfortunately it would not work.
-
it is going to be very similar to this guide
-
yes you can install hexos on mirrored drives on the installer they give you the option
- 1 reply
-
- 2
-
-
PSA: You can update to 24.10.2.2 via TrueNAS; HexOS Q2 Update Imminent
Mobius replied to jonp's topic in Announcements
im not 100% about rolling back but if you do upgrade to fangtooth make sure you don't upgrade zfs pool flags in the truenas ui if you do, you definitely can no longer roll back and still use the same storage pools -
The Mod team has also reached out for you and hopefully we can get this resolved quickly for you
- 1 reply
-
- 1
-
-
But we can expect some more soon
-
Truenas permissions can get real frustrating. Even understanding them is tough. Hopefully we can avoid the tn ui the further Hexos matures
-
i have always just used rsync in the truenas ui shell, does that work for which users you are trying with?
-
PSA: You can update to 24.10.2.2 via TrueNAS; HexOS Q2 Update Imminent
Mobius replied to jonp's topic in Announcements
q2 update has long released and we are most likely closer to the release date of the q3 update than the q2 update -
its all preference
-
yikes, thats super unlucky. verbatim had a reputation for making quality drives so im quite surprised.
-
happy i could help
-
so out of the box it likely would not support both since there is only one full size pcie slot. I think there is an extra x4 pcie slot but it would like likely need some modification to allow larger devices to fit (or you need to purchase an adapter) and even then im not 100% if itll work or not. the 12600k works really well with plex specifically because of the optimizations intel made with intel quick sync. you likley don't need need a gpu.
-
Thinking from swapping from Proxmos.. couple of question before I take the plunge.
Mobius replied to Morphexe's question in OS & Features
I don't have any experience with unraid so I'm wondering what makes it harder to use? -
Thinking from swapping from Proxmos.. couple of question before I take the plunge.
Mobius replied to Morphexe's question in OS & Features
At this time hexos doesn't handle the creating or managing of VMs as far as i am aware, the feature was delayed while main truenas scale figures out what they wanna do. Truenas scale had transitioned to incus in TN scale 25 but they started detransitioning or something While you can setup and use VMs for now it'll have to be done via the truenas interface i haven't done it myself yet so i can't really comment on its ease. At the very least gpu pass through works for me for jellyfin and immich, i haven't tested anything else. Rule of thumb is that if it works on truenas scale it'll most likely work on hexos I have multiple zfs pools that are accessed by multiple mac and windows computers at the same time. Hexos uses the SMB (samba) protocol to make shares available to the network and i do believe most linux distros support smb. at this time there are some hoops to jump through. You need to either setup a reverse proxy, something like tailscale or a cloudflare/pangolin tunnel. i found setting up a cloudflare tunnel was most straight forward but you need to own a domain for that. always happy to help. honestly tho if you are running 20+ vms proxmox makes sense. you can even run hexos as a vm in proxmox and let it handle all your non vm needs. hexos also has a 30 day refund policy so you can try and see if you think it'll work for you. -
we do get updates on what to expect with each blogpost with this one being the latest one https://hexos.com/blog/status-update-amp-zfs-leadership-call-tonight but yeah it would be a good idea if any updates were incorporated into the roadmap page as well. I imagine the q3 update blog post will be arriving near the end of q3
-
the standard for plex/jellyfin is onboard intel igpu because it has access to intel quicksync. as a bonus you will be saving a LOT of money over the years not paying for dgpu idle electricity costs
-
@ubergeek has swapped out his hardware many times iirc. as long as boot drives stay the same, its fairly straight forward swapping parts. realistically its probably the lsi 9300-16i but that one chugs on electricity and outputs a lot of heat so you would need a dedicated fan for it and in the long run itll be more expensive. i personally bought an oem variant of the 9400-16i for not that much more than 9300s are going for on ebay. i forget the model numbers, probably best you just ask chat gpt what the oem model numbers are