tezgno

I was having a lot of problems with performance in my environment until very recently. I couldn't quite put a finger on when I started noticing things getting really bad but it was definitely getting worse over the past year. Then, a couple of days ago, I stumbled upon some things that helped me to resolve my slowness problem on Mac. Essentially, for the past several years the performance of SMB has gotten worse and with most of us trying to tinker with settings to improve performance, which many of us did, it would seem that around last year, limetech actually decided to work on improved settings for SMB... which after looking at this post here, caused me to figure out that all of my settings and changes combined with their settings and changes was actually the culprit. And, interestingly enough, that was around the time that I started to notice performance getting really bad for me. The fix for me was to remove all of my SMB extra settings completely, copy the fruit conf file that is in the post, add my appropriate changes that I wanted to be default, removed the ones that they explicitly state shouldn't be working, and restarted samba. When I tell you performance was night and day, it would be an understatement. To confirm that was the issue, I undid everything and put it back to what it was before and sure enough, performance tanked. To put into perspective, what would take only a few seconds to copy on my 10gb network was taking hours. Now, things are back to taking seconds. My advice would be that if you have made your own settings changes to unRAID over the years in order to improve performance, especially in the smb extras files, if you are running 6.11 or newer, you may want to re-evaluate your settings against the default, as you may be inadvertently negating options designed to improve performance from some of the changes put in. That is what was happening to myself.

Ok, so I figured it out what was going on. There are several things that are needed in order to make it work with Oauth2. In my case, I'm using Nginx Proxy Manager: In the config.php, you need to add your trusted domain and trusted proxy IP address. Trusted proxy is an array and you put the IP address of your proxy. In the default file inside site-confs (nginx folder), you need to add a rewrite: # The following rule is added for Oath Configuration from a reverse proxy rewrite ^/.well-known/openid-configuration /index.php/apps/openidconnect/config permanent; You have to restart the container after adding the config. Finally, in proxy manager do not include any exploit blocking (turn off Block Common Exploits). After doing these changes, I can now use proxy manager both on the web and the desktop clients work.

I have the same problem. I think this has something to do with OAuth2 and using a reverse proxy. When I change the URL to the IP address of the container, I get the login screen. However, when I use the domain from the reverse proxy, I get a 403 error. I noticed that there is a configuration missing in nginx (for /.well-known/openid-configuration). But I think there is something else missing that is preventing this. If I figure it out I'll post it.

I did give something similar a try to and it was not able to work the minute I select /mnt/user. I suspect that it may have been permissions related, but I'm not sure there either. Duplicacy thus far has worked really well. I used to use Duplicati, but the failure rate of it was too much to bare. Thus far, Duplicacy has worked really well. I do find that it produces slightly larger backups, but I feel that much of that can be controlled with higher compression settings if it became an issue.

No, I was not able to get it resolved. I decided to use Duplicacy. It's cheaper, doesn't have the same limitations, and thus far, has worked (even did a restore with it and it worked fine).

Just an update, the upgrade didn't solve the problem. However, further digging into the issue yielded an issue being caused by Duplicati (Docker Container). Looks like the choice by Duplicati of using a database was the ultimate culprit. Switched from Duplicati to another backup solution (Duplicacy in this case) and system has been stable ever since.

Sure, As you can see, storage is completely missing even though it is mapped to /mnt/user. If I change it to anything other than /mnt/user, it shows up.

I just installed this docker with the default settings. For some reason, the /storage does not show up under the GUI for me to select. I know that it is mounted because when I go to create a backup target and choose filesystem, I can clearly see that /storage is there and points to /mnt/user. However, it doesn't show in the GUI. If I change it from /mnt/user to, say, /mnt, it does show up, but user is missing (it shows cache and disk, etc, but not user). Anyone else experiencing that?

Just pulled down the 6.10 rc1 upgrade and switched over to ipvlan. This should be easy to test because if it stays running for longer than 24 hours, that may have solved the issue for me. Fingers crossed.

Hello everyone, I've been trying to figure out the root cause to some recent system instability that seemed to have began about 3-4 weeks ago. Almost daily, I'm having to reboot my system due to high CPU usage or complete crash. After rebooting, system runs fine for about 24 hours and then crashing again. Not sure if this is hardware related or something else. I have an AMD system and did disable c-states. Prior to recently, everything worked fine with no issues. I have attached the diagnostics from today as I was able to capture the diagnostic from this mornings reboot up until crash. Any help would be appreciated. plexbeast-diagnostics-20210907-1613.zip

I, too, just started using unRAID myself. Been following it for sometime but decided to make the switch since my Synology is nearly out of support and due a replacement. So far, it's a night and day difference. Naturally, the Synology is more simplistic and user friendly, but unRAID is way more powerful. So far, really liking the decision.