John_M

I haven't fully thought this idea through or done any testing but suppose you were to create a user share that contains your preferred folder structure but don't actually enable sharing on it. Then create a user share for each user (x, y, z in your example) that contains symlinks to the actual files you want that particular user to be able to access. I'm not sure whether it would work and I'm happy for someone to shoot the idea down in flames. It seems your problem needs a bit of lateral thought so I'm just making a suggestion.

Forgive me for stating the obvious, but did you plug the USB cable in?

I re-installed the plugin and tested it yesterday. It seems to be working well again with none of the syslog spam that older versions caused. Many thanks for your time and effort.

Since a file's name is separate from its content I would expect nothing to be done when the name is changed. The content remains unaltered; therefore so does the hash. The name is part of the file's metadata and editing it actually affects the parent directory. So if directories were hashed as well as files, then it's the directory that contains the file that would have to be rehashed.

The question was asked a couple of years ago: I guess @RobJ's advice still applies.

I'm seeing a lot of this in the syslog of one server that uses https for the web GUI: Aug 22 12:39:40 Northolt nginx: 2018/08/22 12:39:40 [error] 7259#7259: OCSP responder timed out (110: Connection timed out) while requesting certificate status, responder: ocsp.int-x3.letsencrypt.org, peer: 2.21.67.65:80, certificate: "/etc/ssl/certs/unraid_bundle.pem" If I understand it correctly, the problem is outside my network. It happens every fifteen minutes or so and it's been happening continuously since 19 August (three days ago). It looks as though it's trying to verify the authenticity of the certificate but getting no response from LetsEncrypt's server. The current certificate is valid until 4 November. A second server, also using the https://#hash.unraid.net method isn't showing the error but it's certificate is due to expire on 28 October. Is there anything I can do or should I just wait? I'm considering reverting to http. northolt-diagnostics-20180822-1329.zip

Two things I can think of that might bloat a thinly provisioned image. Has the virtual disk become badly fragmented? Does it use CoW?

Is this plugin being maintained again? Is it now compatible with unRAID 6.5.3? I ask because I had to uninstall it and wondered if it's safe to use again.

From your description is looks as though the text-based part of the boot sequence is displaying normally but the problem happens when the output switches to graphical mode. Try the nomodeset tweak here:

There is a subtle difference between /mnt/user/appdata and /mnt/cache/appdata even when the share is set to cache only. The former is handled by shfs while the latter isn't. Hence Squid's suggestion to bypass shfs. I thought this issue had been fixed in the meantime, though.

I haven't seen it mentioned in the release notes, though I wasn't particularly looking for it. Have you tried looking there?

There are lots of useful goodies in the diagnostics! That one is in system/motherboard.txt

ASUSTeK COMPUTER INC. - CROSSHAIR V FORMULA-Z BIOS Information Vendor: American Megatrends Inc. Version: 1602 Release Date: 09/16/2013 I had the advantage of having to look at your diagnostics to find out

Why don't you update your BIOS? You should also update unRAID. Version 6.5.2 is available and has a newer kernel.

Tom has said that unRAID 6.6 will use the 4.16 kernel. It looks as though 4.15 will be skipped as 4.16.8 is already out. In anticipation of the obvious follow up question, I really don't know when.

I don't have one myself but I would like to see more HBAs supported since the Marvell-based ones are proving to be so troublesome for many people nowadays.

The k10temp module has apparently been modified to include Zen and is included with the 4.15 kernel. See here. Motherboard temperatures and fan speeds rely on whatever ITE, Nuvoton, etc. chip is used and whether it's detected and supported.

Do you have it connected to a USB 3.0 (USB 3.1 Gen 1) port? Do you have a USB 2.0 port that you can try?

It's like he said: root@Mandaue:~# ls -l /sbin/reboot lrwxrwxrwx 1 root root 4 Apr 21 19:00 /sbin/reboot -> halt* root@Mandaue:~# ls -l /usr/local/sbin/reboot /bin/ls: cannot access '/usr/local/sbin/reboot': No such file or directory root@Mandaue:~#

I notice the sg_format command isn't included in unRAID. Maybe it could be added to the NerdPack.

What happens when you click the Mount button?

OSXFuse is available as a .dmg from here. Then you need the Fuse-XFS module from here, also a .dmg, so you won't have to compile anything from source. Note that it works in read-only mode - writes are not implemented - and the journal is ignored. The XFS module is alpha software and hasn't been updated for a couple of years but I used it a year or so ago and it worked on Mountain Lion. The disk won't automount but see the Readme that's included: fuse-xfs /dev/rdisk1s1 -- /mnt/xfs though the standard macOS mount point is below /Volumes rather than /mnt.

What version did you upgrade from? The reason I ask is that I'm currently using 6.5.1-rc6 and I don't see the problem you report so I'm wondering if the cause is in the diff between -rc6 and -final.

If it's repeatable you might well have found a bug (and its workaround!). I suggest you post the details in a new thread in the pre-release section. @bonienl will be pleased to hear about it.

I'd run find /mnt/user -name @eaDir -exec rm '{}' \; but because it's potentially dangerous I'd run find /mnt/user -name @eaDir -exec ls '{}' \; first to find and list them.