Been using RN314 for over a year, It has been working fine until this update. Noted that users were having issues with shares previously mounted after a few days, or could not mount them again after reboot. I was using and Underscore in the Device Name "ReadyNas_314" and after the upgrade the network could not be found any more, it seems it shortened it to just "ReadyNas". When I tried to remove the underscore and shorten the name a few days later when we could no longer mount shares at all, it would change in the frontview but return to old name after reboot. I logged in using ssh and found the hostname was correct but the hosts file still had the old name with the underscore. When I tried to vi the hosts file, got message that there was no "swap" available so I could not save any changes.
I checked the message log and found that there were errors rotating the logs and said it was out of space, I could not find any mounted drives that were anywhere near even 50% full, but I deleted two of the log files to make space and rebooted the unit from the command line.
After a reboot it seemed that the naming issue was resolved and I could vi the hosts file and fix that minor error, but we still had problems connecting to the shares.
I had created several groups of users and we have used them to access shares that had the group name assigned for access but, after installing 6..9.2 firmware update, users could no longer access shares that belonged to the group they were assigned to and I had to remove the group access to shares and assign access to each individual user.
I managed to get it all working again but it took me days to troubleshoot and get the unit to opeartional status again. I somebody else might find the information useful and of course maybe tech support can fix the firmware.
I understand the "Underscore" issue but I have been using it in Unix based systems for about 30 years and It worked fine in the previous firmware 6.9.1 and there was no mention of the changed behavior in the change notes. It would have been nice to know before implementing the new version.
Also that does not explain the other issue of the "swap" space crash and log space rotation issue.
And the fact that group access no longer works after the update, All the group access information seems to be intact and no signs of corruption, it just stopped working.
I understand the "Underscore" issue but I have been using it in Unix based systems for about 30 years and It worked fine in the previous firmware 6.9.1
Actually Netgear started enforcing the RFC character limits last April (in 6.7.1). Are you sure you upgraded from 6.9.1?
I agree it should have been in the release notes (and wasn't).
wrote:
Also that does not explain the other issue of the "swap" space crash and log space rotation issue.
And the fact that group access no longer works after the update, All the group access information seems to be intact and no signs of corruption, it just stopped working.
Understood. I think you'll need some analysis from Netgear on those.
I have been unable to find a way to report the issue of Groups no longer working on 6.9.2 firmware update. Do you have to purchase support to report an issue???
When the problems started a few days after the upgrade from 6.9.1 to 6.9.2, and yes I am sure of the previous version. I tried a few time to rename the ReadyNAS by removing the underscore. Each time I made the change and rebooted it after confirming the change was made, it reverted back to the original name after the restart. Thats when I realized something was wrong and I used an ssh login to start looking at what problems were occuring and found that there were the errors described in my original post.
The group logins that were failing are all using SMB and running windows 10, I looked at the logs thru an ssh login while trying to login with a group account and it showed a successful login for the user account and then a failure to login to mount the group share.
The only changes made were to add share access for each individual account to be added to each share since the group logins no longer worked.
Configuration changes that don't stick are often a symptom of a full root partition - that is why I asked about it. If you used df or du to assess the fullness, you might also try using the btrfs commands (which include metadata that df and du don't see). Worth checking, just in case - but probably not likely.
I don't quite get what is significantly different from the df command data, maybe it tells you something more. I will however keep in in mind for any future troubleshooting.
