NLS

I hope you raised a support ticket or something.

It is possible that after some time-out the system checks the secondary DNS, that works. I am not sure how Linux does switch to secondary DNS for resolution and if it is configurable. Since people hosting DNS in the server itself is not uncommon (I mean... it is the server), you should probably consider alternative solutions. Can you still make the checks and log issues, but not freeze the system? (make it a background task with some time out) Now, if those FQDN are NEEDED during boot (and is not just a check), you should also have them as static entries and/or temporarily use an extrernal DNS (like 8.8.8.8), that will also check connectivity to Internet. Maybe both (try defined DNS in background, but test with alternative DNS and on time-out use static entries). Of course it can also revert to the check later after server boots and anything auto-start is working. (so maybe raise a warning flag and then clear it?)

For #1. But is it? I mean BTRFS handles bit-rot and (about the tolerance to external issues) is CoW. For #2. That is GREAT news and seems I didn't notice in the latest release notes. I hope they are actively working on this. Thanks for you reply, I still need actual users that have tried BTRFS on array and even better, compression, to chime in. I am 100% sure we can find at least one.

I would like to (re)open this discussion, hopefully people that have already done it will chime in and tell us what is positive and negative in this move. So to organize this a bit: 1) Official docs, claim that BTRFS is not as robust as XFS yet. But they use BTRFS by default in cache (where the risk is considered smaller? Yet most people's VMs are there). Is this claim still valid in Oct 22? Or should UNRAID move to default BTRFS "soonish"? 2) Compression, while default in most "competing" solutions that use ZFS, is not even a (GUI) option in UNRAID. People that use it (I believe there are a few), feel it should be an option that can be set "officially"? (or even default) 3) Recovery options when moving to BTRFS or moving to BTRFS with compression. Are they available? Is it the same as with XFS? 4) Load on CPU? Actual saving benefits? Some people claim that for certain uses, writing with compression (or reading compressed) can even speed up things, as less data are read from the actual slow media (or writen to it) and the load moves to the CPU. In practice? 5) For people that have actually done it: Can someone give a streamlined stupid-proof process on how to achieve BTRFS with compression in UNRAID?

Same here.

Can I resurrect this thread and ask how this ended up? Did it work properly? Did the performance improve or became worse? Other gains or losses?

Well DNS is on a VM inside UNRAID, so it is not there while booting. The secondary one (in UNRAID settings) is my router, so that should work. So... if that is the issue probably it should skip to secondary DNS much faster OR use maybe a hardcoded "boot time" DNS or (third possible solution) revert to a (list of?) hard-coded name resolution entries (which is the best option probably). It was working faster somehow before a few weeks probably (not sure, I rarely reboot).

As I said, I removed that addin driver when I changed motherboard. Anyway, here are my diags... quasar-ultima-diagnostics-20221010-0101.zip

I am talking maybe even 5 minutes waiting. Even though I know that it takes time, it is so much, that I start thinking every time that something is wrong this time. And then it proceeds. Here is where the long pause takes place: Also, I am not sure why it looks for nct6775 (don't know if it is related to the long pause), as I don't have that plugin in UNRAID (it was in my previous motherboard, but I removed it and I have ITE driver now). But I don't know if it comes from that removed plugin either. So: - Any idea how to find what makes the huge pause? - Any idea how to remove this modprobe for nct6775? (or is it part of the boot process anyway?)

Tried autofan. Doesn't detect anything. My mobo is GB X570 UD. Temp plugin works fine. Any ideas?

Since you mentioned sleep... I was playing around with this plugin. Server sleeps fine. BUT HOW do I wake it up "transparently"? Where transparently means "on access" basically. When I try to access a share for example. I know how to send magic packet etc. but I want this to work transparently, not manually. Is there a way?

hah... we wrote the same thing the same time

what are you talking about?

Just a note. I made Geyser work with this server after all. It always showed as working (when server started it reported geyser loading and listening to 0.0.0.0:25575 - that was ok from the beginning), but in reality it was unreachable (from own LAN). Don't know what made it work but here are some things I tried: - Made login mode floodgate (also have floodgate). It shouldn't be this, because my issue was that server was not even responding from bedrock client. I did have to whitelist the player when game worked (it didn't show with same UUID). - Made an extra port setting in the container where I also "opened" UDP ports for Geyser (as I noticed it kept mentioning about UDP port). It could be it. ...some other things I don't remember (not a very useful port eh? )...

It's pretty clear that it was a very useful plugin. Hope it somehow survives.

#2. They were installed using CA. I am pretty positive I didn't use a manual compose or anything. Remember CA DOES update them (or UNRAID does it?), even though it doesn't show them as installed. Is there a way to "force" CA to "see" them? (they are in the same place that CA puts them anyway) Is there a file to check or edit or something?

@Squid hello, two things to report I just noticed 1) I tried to update two containers right from CA interface (instead of UNRAID dashboard I always use). CA reported that they needed update so I tried through CA just to check it out. Well... both failed. It opened a white window where it listed the process/progress, but it failed to build the first container, then failed the second. It continued being "blue" (even after page refresh) in dashboard, so I updated from "docker" tab. This time it didn't load any new packages (signifying that probably through CA it *did* get them), but the container command was successful this time. It might just been some hiccup, I will try it again with next update I see. 2) I just found out that two of the containers I use (maybe more actually, but certainly just few), although they are installed AND used AND even updated (one was even updated today), they don't report as "installed" in CA. The two I found were from linuxserver's repo, but this could just be by chance, as other containers from that repo show as installed properly. One that doesn't show as installed is qbittorrent, other is syncthing. Not easy to re-install any of the two (scared to lose their config). (please tag me as I cannot follow this huge thread)

Yes please. I configured it (I think) - but it never worked.

I have no graphics in the stats. Even after I applied this:

Hope this is fixed soon.

I edited my OP just to add an "S" to "year" (that I use UNRAID). I suspect 1 year vs 15+ (with a break in between) must be different. (including how much noob, people consider you when replying )

There are very nice GUI based duplicate finders in the apps.

To be honest, I have already tried to ignore possible risks in the yearS I use UNRAID and use my server more or less normally. BUT I would like to have some expert (or official) feedback on how UNRAID handles a file write (or many files write), while data are rebuilding on a disk. And before people saying that writing usually involves the cache, because it is not always the case... (a) people without cache, (b) shares that are "no cache", (c) file deletes (that also affect the contents of the disk). More specifically, how does it handle writing: - to a functional HD in disk blocks that the data rebuild has already passed over. - to a functional HD in disk blocks that the data rebuild hasn't yet passed over. - to the emulated (and data reconstructing) HD in disk blocks that the data rebuild has already passed over. - to the emulated (and data reconstructing) HD in disk blocks that the data rebuild hasn't yet passed over. ...and the rare case that the filesystem tries to use blocks on functional or emulated HD... DURING the passing of the rebuild from those very disk blocks. I want to assume that UNRAID handles all those cases gracefully somehow. But is it the case or someone plays with fire when writing to an array while rebuilding?

Yes that is what I thought, but I open the XML and I cannot find that "device". That's the issue. I actually created a new Win11 VM (just to make a fresh XML) and the only difference between the two XML is this: (all other lines are same or have reasonable differences that I can explain) Old (with problem) <memballoon model='virtio'> <address type='pci' domain='0x0000' bus='0x04' slot='0x00' function='0x0'/> </memballoon> New (haven't tested though) <hostdev mode='subsystem' type='pci' managed='yes'> <driver name='vfio'/> <source> <address domain='0x0000' bus='0x0a' slot='0x00' function='0x1'/> </source> <address type='pci' domain='0x0000' bus='0x04' slot='0x00' function='0x0'/> </hostdev> <memballoon model='virtio'> <address type='pci' domain='0x0000' bus='0x05' slot='0x00' function='0x0'/> </memballoon> So is it the different bus of VirtIO? Is it the extra "subsystem" that the new VM reports (and added to XML)? EDIT: WAIT! It is the other way around! The problem VM has the extra lines! So it probably is that. I will try. EDIT #2: Seems that was it. I am not sure what I lost by removing vfio subsystem...

I will try the first. I think this mobo has an ITE chip. I installed it but doesn't detect anything. Maybe reboot the system first? The other (NCT6687), although I didn't have explicitly installed (as a plugin), it was already there probably because of my previous mobo. But I don't see it in my plugins, how can the system try to use? Even when booting UNRAID, I see in the boot scroll that "nct6687" fails to load at some point. Why it tries to use it, when I don't have it installed? I even pressed "unload drivers" in the plugin.