[Support] Nginx Proxy Manager (NPM) Official

[st1994]

4 hours ago, Kilrah said:

NPM made some changes that break things, there are already issues open about it on github, several of us are staying on version 2.9.22 until they decide how they're going to handle it...

 

Thanks, i'll look into downgrading my container version.

[tommykmusic]

I am trying to setup nextcloud with a cloudflare tunnel using nginx and I got everything setup following ibras video. My only issue is that I get a 502 error when trying to access nextcloud outside of my network.

 

Upon checking my nginx log I see this error

[app         ] Duplicate relation "access_list" in a relation expression. You should use "a.[b, c]" instead of "[a.b, a.c]". This will cause an error in objection 2.0

 

as well as this in blue

[app         ] [4/15/2023] [4:09:24 AM] [Migrate  ] › ℹ  info      Current database version: none

 

I can't seem to figure out what could be causing the issue as I don't have any access lists. 

Any help would be appreciated, Thank you.

Edited April 15, 2023 by tommykmusic
code fix

[profzelonka]

On 2/22/2023 at 1:51 AM, Kilrah said:

Likely you didn't add the trusted proxies entry in HA configuration:

 

 

Where can I add this trusted proxies on a VM that's not for HomeAssistant? My VM houses diaspora a decentralized social network - other than that I have the exact same setup as this post. Getting error 502 bad gateway.

 

Can this be set in nginx somewhere on the VM?

[mgutt]

5 hours ago, profzelonka said:

Getting error 502 bad gateway.

Read the first page. br0 can't reach bridge networks as long you don't enable the tweak in the docker settings. Enable this or change npm's network to br0, too.

 

 

[profzelonka]

1 hour ago, mgutt said:

Read the first page. br0 can't reach bridge networks as long you don't enable the tweak in the docker settings. Enable this or change npm's network to br0, too.

Hey thanks! I changed docker setting's "Host access to custom networks" to Enabled. Still getting a 502 error though. Do you mean to change the npm docker's network to br0? Cause I'm using it for other dockers already so I guess that won't work for me. Is there anything else I can try to resolve my 502 error? I can access the VM via IP but not thru the proxy still.

[mgutt]

8 hours ago, profzelonka said:

Is there anything else I can try to resolve my 502 error?

Yes, reading the first page.

[profzelonka]

On 8/1/2021 at 1:10 PM, mgutt said:

My suggestion: Run NPM as host and run all other containers as bridge.

Is this what you're trying to refer to? Went ahead and set NPM as host - webui doesn't load anymore, switched it back to bridge, now NPM is a complete dud, no webui and for some reason the port reverted to 8181 from a custom one I set. Stopped it, set it back to my port, still can't access NPM any more..

 

Come on man.. it's 5 am for me, just gonna call it a failed no sleep night I guess.

[Kilrah]

To run the NPM container as host you need to change the ports the unraid UI runs on in settings->management access so 80 and 443 are free for NPM.

[profzelonka]

I have NPM running on 1443 and 180 instead, with router directing web traffic to those unique NPM ports. But now NPM is completely not working since I set it to host, changing it back to bridge it shows up as running but the webserver and webui are inaccessible. /facepalm

Edited April 16, 2023 by profzelonka

[mgutt]

4 minutes ago, profzelonka said:

I have NPM running on 1443 and 180 instead

Ports can't be "changed". You are only forwarding port 1443 to Port 443 and this is only possible in the bridge network. With host or br0 a container uses it's default ports and those are 80 and 443 for NPM.

 

Personally I prefer changing unraids ports and run NPM on host network with port 80 and 443, because this is the only reliable solution for IPv6.

[Kilrah]

14 minutes ago, profzelonka said:

I have NPM running on 1443 and 180 instead, with router directing web traffic to those unique NPM ports.

You cannot run on host or br0 with changed ports, port mappings are only possible on bridge.

 

If you tried changing them while in host mode you need to set them back. 

 

  

8 minutes ago, mgutt said:

Personally I prefer changing unraids ports and run NPM on host network with port 80 and 443

Same.

Edited April 16, 2023 by Kilrah

[profzelonka]

Set unraid ports to something else. Removed NPM, readded, still can't use it.. Looks like old ports are being remembered somewhere. How do I wipe it?

[Kilrah]

If you did "reinstall from previous" that would have kept your previous settings. 

 

Either install clean or just check they're set right.

 

[st1994]

On 4/13/2023 at 4:51 AM, Kilrah said:

NPM made some changes that break things, there are already issues open about it on github, several of us are staying on version 2.9.22 until they decide how they're going to handle it...

Sadly this didn't fix the issue.  Still having the same problem where NPM doesn't start properly after i back up the container nightly.

[aglyons]

regarding the "because parent directory has insecure permissions" errors that were discussed in earlier threads.

 

As I understand it, NPM is looking at the permissions of the parent folder. That folder is in the appdata folder on UR. UR wants the ownership and permissions of its contents as 'nobody' and 777. There's even the docker permissions tool that fixes this in one click.

 

But if UR sets the permissions to what it needs, NPM will complain that this is insecure. 

 

How can we get around this problem?

 

I'm sorry if this is a rehash and has already been resolved. I've spent that last 30 minutes searching the web for answers with no luck.

Edited April 22, 2023 by aglyons

[flex]

Running the latest version of NPM. The container is writing to the docker img location. It's writing/deleting as it fluctuates, at times filling with docker img completely. It is not the logs writing to the img. Whatever its writing clears the space upon the container stopping. It is writing to /etc/hosts as shown below.

shfs            932G  644G  281G  70% /data
/dev/loop2       20G   16G  3.6G  82% /etc/hosts

[aglyons]

17 hours ago, flex said:

Running the latest version of NPM. The container is writing to the docker img location. It's writing/deleting as it fluctuates, at times filling with docker img completely. It is not the logs writing to the img. Whatever its writing clears the space upon the container stopping. It is writing to /etc/hosts as shown below.

shfs            932G  644G  281G  70% /data
/dev/loop2       20G   16G  3.6G  82% /etc/hosts

 

On the advice of someone else, I switched from using docker.img to a DIR. It's more flexible with how much data can be written as the IMG approach has a fixed virtual disk size. THe DIR route is the capacity of the share it's being stored in. You can chagne that in the Docker settings after you shut down Docker. You will have to reinstall all your containers but that's fairly simple as they are all listed in 'Previous Apps' in the CA.

 

[flex]

28 minutes ago, aglyons said:

 

On the advice of someone else, I switched from using docker.img to a DIR. It's more flexible with how much data can be written as the IMG approach has a fixed virtual disk size. THe DIR route is the capacity of the share it's being stored in. You can chagne that in the Docker settings after you shut down Docker. You will have to reinstall all your containers but that's fairly simple as they are all listed in 'Previous Apps' in the CA.

 

So instead of filling up the docker img its now going to fill up my SSD. Doesnt seem like a proper solution. What I would like to know is what kind of data is even being written here where it's writing over 15GB within a few minutes and why you cannot change where the data is written to.

[aglyons]

Just now, flex said:

So instead of filling up the docker img its now going to fill up my SSD. Doesnt seem like a proper solution. What I would like to know is what kind of data is even being written here where it's writing over 15GB within a few minutes and why you cannot change where the data is written to.

 

Setting your IMG to 20GB and having it fill up can crash every container you have running. A 2tb ssd will take a heck of a lot longer to fill up and you would probably notice this before any damage happens. But you do you.

 

Quote

What I would like to know is what kind of data is even being written here where it's writing over 15GB within a few minutes and why you cannot change where the data is written to.

 

Best bet would be to file a post on the Github for the project. - https://github.com/NginxProxyManager/nginx-proxy-manager/issues

 

 

[flex]

1 hour ago, aglyons said:

Setting your IMG to 20GB and having it fill up can crash every container you have running. A 2tb ssd will take a heck of a lot longer to fill up and you would probably notice this before any damage happens. But you do you.

 

You assume that the drive is empty. it is not, moreso I would prefer not to have unnecessary wear on a main drive. Even though it is mirrored, the less stress it goes through the better.

I'm not rejecting your comment but it's a bit like having a leaky faucet in which you have a bucket under that you empty ever so often and instead of fixing the leak you just put a bigger container so you have to empty it out less often.

[aglyons]

1 minute ago, flex said:

 

You assume that the drive is empty. it is not, moreso I would prefer not to have unnecessary wear on a main drive. Even though it is mirrored, the less stress it goes through the better.

I'm not rejecting your comment but it's a bit like having a leaky faucet in which you have a bucket under that you empty ever so often and instead of fixing the leak you just put a bigger container so you have to empty it out less often.

 

I figured "But you do you." would be the end of that convo.

[Herobrine]

For a while now whenever I try to add/edit sites from the dashboard I get a internal errors till I restart the container. At most I am able to successfully create 1 new proxy host before I need to restart the container after making the config change (which sends a internal error warning) then making me restart eh container to apply it. 

[fir3drag0n]

What's the latest update on the newest nginx proxy manager update? Is it still buggy and recommended to stay at an older version or can I upgrade? And if yes, how?

[fir3drag0n]

On 4/26/2023 at 5:15 PM, Herobrine said:

For a while now whenever I try to add/edit sites from the dashboard I get a internal errors till I restart the container. At most I am able to successfully create 1 new proxy host before I need to restart the container after making the config change (which sends a internal error warning) then making me restart eh container to apply it. 

Same problem 

[MAM59]

7 hours ago, fir3drag0n said:

? And if yes, how?

Dunno current status, I stay with the old version (and am fine with it).

 

How? easy!

go to the docker's settings page and locate the "source" entry:

(in german "Quelle")

At the end of the line you will have ":latests", change this to ":2.9.22" and hit ok. It will downgrade to this version and never ask for an update again.

(of course you can later on decide to put back :latest or any other version number you like to run)