SOLVED Jails accessing different version of mounted storage from rest of system?

[Mikrz]

So, I'm pulling my hair out over this. My software running within these jails is completely choked up on this as am I. As far as I can tell, I am currently accessing two different sections of the filesystem despite all the paths pointing to the same spot.

To give a breakdown of my situation:

• I have 3 different "systems" including jails that are accessing this mount point.
• I have Deluge, Sonarr, and my Windows machine through an SMB share.
• All has been working until the past 4 or 5 days, after the first automatic drive scrub occurred since I initially setup the system.
• Deluge and Sonarr, after remounting the "Media" dataset in each, see the same thing. If I touch test in the Deluge jail shell, the test file shows up in Sonarr jail shell.
• If I try to view the exact same dataset from the SMB share on my windows machine, there are folders missing and the test file does not appear. The same is the case if I SSH directly into FreeNAS as my user, or as root via su.
• If I create a test file in the FreeNAS shell, or the windows share, they can see that file but the Jails cannot. I have attempted rebooting, to no avail. I have tried restarting the SMB service, that didn't help either. I have ensured all permissions are accurate, and for testing as permissive as possible.

I am completely at a loss here.

FreeNAS Nightly version: Latest as of 9pm EST September 11th 2017 (attempted an update to see if that would remedy the situation).
16gb of ram, Intel i3-6100, 2x4TB WD Red drives in STRIPE.

 

[dlavigne]

If you reboot into the STABLE train and attempt a test, does anything show up in /var/log/messages?

 

[Mikrz]

My apologies, it has been quite a while since you replied. A lot came up in life. As it turns out, somehow the WebUI failed to update the displayed text (or was somehow cached, I don't know) and the jails were mounting to one point which was an older dataset I was switching from, while my Windows share was correctly mounting to the new dataset location. Seems this was more of a PEBKAC issue as I didn't think to test on a system other than my own desktop. I appreciate your help nonetheless dlavigne.