[GH-ISSUE #727] "Bad Gateway" at Login Page after Upgrading to 2.7.0 #615

Closed
opened 2026-02-26 06:33:39 +03:00 by kerem · 21 comments
Owner

Originally created by @Vocta1310 on GitHub (Nov 18, 2020).
Original GitHub issue: https://github.com/NginxProxyManager/nginx-proxy-manager/issues/727

Hi, I can't login anymore in configuration Panel after Upgrading to 2.7.0.
Before the Upgrade everything works great.

What can I do to fix this issue?

Originally created by @Vocta1310 on GitHub (Nov 18, 2020). Original GitHub issue: https://github.com/NginxProxyManager/nginx-proxy-manager/issues/727 Hi, I can't login anymore in configuration Panel after Upgrading to 2.7.0. Before the Upgrade everything works great. What can I do to fix this issue?
kerem closed this issue 2026-02-26 06:33:40 +03:00
Author
Owner

@jmeinicke commented on GitHub (Nov 18, 2020):

my solution was to add the database env to my compose file and remove the config.json config file from mount into container

https://nginxproxymanager.com/setup/

<!-- gh-comment-id:729541351 --> @jmeinicke commented on GitHub (Nov 18, 2020): my solution was to add the database env to my compose file and remove the config.json config file from mount into container https://nginxproxymanager.com/setup/
Author
Owner

@d33pjs commented on GitHub (Nov 18, 2020):

Having the exact same Problem.

I'm using SQLite. Are there anywhere examples what database environment variables I have to add to my compose file and what removing the config.json fiel from mount into container means?

<!-- gh-comment-id:729556140 --> @d33pjs commented on GitHub (Nov 18, 2020): Having the exact same Problem. I'm using SQLite. Are there anywhere examples what database environment variables I have to add to my compose file and what removing the config.json fiel from mount into container means?
Author
Owner

@Vocta1310 commented on GitHub (Nov 18, 2020):

Fixed with 2.7.1 - Thank You @jc21

<!-- gh-comment-id:729709160 --> @Vocta1310 commented on GitHub (Nov 18, 2020): Fixed with 2.7.1 - Thank You @jc21
Author
Owner

@d33pjs commented on GitHub (Nov 18, 2020):

Fixed also on my side with 2.7.1 :-). Thank you @jc21 .

<!-- gh-comment-id:729722802 --> @d33pjs commented on GitHub (Nov 18, 2020): Fixed also on my side with 2.7.1 :-). Thank you @jc21 .
Author
Owner

@2Ben commented on GitHub (Nov 18, 2020):

Seems not fixed for me on 2.7.1. I just did a docker pull to upgrade, should I do something else?
Edit: sorry it is actually fixed. Thanks @jc21 !

<!-- gh-comment-id:729776731 --> @2Ben commented on GitHub (Nov 18, 2020): Seems not fixed for me on 2.7.1. I just did a docker pull to upgrade, should I do something else? Edit: sorry it is actually fixed. Thanks @jc21 !
Author
Owner

@stavros-k commented on GitHub (Nov 18, 2020):

Does the container auto updates? I pulled new image, restarted the container but i'm still on 2.6.2

<!-- gh-comment-id:729790178 --> @stavros-k commented on GitHub (Nov 18, 2020): Does the container auto updates? I pulled new image, restarted the container but i'm still on 2.6.2
Author
Owner

@Vocta1310 commented on GitHub (Nov 18, 2020):

@stavros-k - have you cleared your Browser Cache? Maybe that wrong information is stored in the cache ;)

<!-- gh-comment-id:729932581 --> @Vocta1310 commented on GitHub (Nov 18, 2020): @stavros-k - have you cleared your Browser Cache? Maybe that wrong information is stored in the cache ;)
Author
Owner

@stavros-k commented on GitHub (Nov 18, 2020):

@stavros-k - have you cleared your Browser Cache? Maybe that wrong information is stored in the cache ;)

Didn't think of that. But i just tried from incognito and still says 2.6.2

<!-- gh-comment-id:729933220 --> @stavros-k commented on GitHub (Nov 18, 2020): > @stavros-k - have you cleared your Browser Cache? Maybe that wrong information is stored in the cache ;) Didn't think of that. But i just tried from incognito and still says 2.6.2
Author
Owner

@d33pjs commented on GitHub (Nov 19, 2020):

Didn't think of that. But i just tried from incognito and still says 2.6.2

I don't know if that was the question, but here a little insight of what I'm doing to update the nginx-proxy-manager docker container:

docker pull jc21/nginx-proxy-manager:latest
docker stop <id or name>
docker-compose up -d

Afterwards you can remove (all) old images with docker image prune -f

That's it. At least, if you're using docker-compose and if you're using "image: jc21/nginx-proxy-manager:latest" in it.

Hopefully that helps.

<!-- gh-comment-id:730179445 --> @d33pjs commented on GitHub (Nov 19, 2020): > Didn't think of that. But i just tried from incognito and still says 2.6.2 I don't know if that was the question, but here a little insight of what I'm doing to update the nginx-proxy-manager docker container: ``` docker pull jc21/nginx-proxy-manager:latest docker stop <id or name> docker-compose up -d ``` Afterwards you can remove (all) old images with `docker image prune -f` That's it. At least, if you're using docker-compose and if you're using "`image: jc21/nginx-proxy-manager:latest`" in it. Hopefully that helps.
Author
Owner

@stavros-k commented on GitHub (Nov 19, 2020):

Didn't think of that. But i just tried from incognito and still says 2.6.2

I don't know if that was the question, but here a little insight of what I'm doing to update the nginx-proxy-manager docker container:

docker pull jc21/nginx-proxy-manager:latest
docker stop <id or name>
docker-compose up -d

Afterwards you can remove (all) old images with docker image prune -f

That's it. At least, if you're using docker-compose and if you're using "image: jc21/nginx-proxy-manager:latest" in it.

Hopefully that helps.

I'm on Unraid, so most of it its through GUI, but
i removed the docker (its like docker rm)
i did docker imager prune
re"installed" docker (its like docker run) (as repository i have jc21/nginx-proxy-manager but on docker run it says its pulling from :latest)
It didn't find available image so it pulled ALL layers from scratch.

On WebGUI it still says 2.6.2. Is there some way to check from console?
Maybe something in my config isn't compatible with 2.7.1 yet and this keeps it from upgrading?

Just made a new container on different ports and new config mapping, still 2.6.2.
I should probably start new issue , because i don't really have something that doesn't work, and its not relevant with this topic.

<!-- gh-comment-id:730270712 --> @stavros-k commented on GitHub (Nov 19, 2020): > > Didn't think of that. But i just tried from incognito and still says 2.6.2 > > I don't know if that was the question, but here a little insight of what I'm doing to update the nginx-proxy-manager docker container: > > ``` > docker pull jc21/nginx-proxy-manager:latest > docker stop <id or name> > docker-compose up -d > ``` > > Afterwards you can remove (all) old images with `docker image prune -f` > > That's it. At least, if you're using docker-compose and if you're using "`image: jc21/nginx-proxy-manager:latest`" in it. > > Hopefully that helps. I'm on Unraid, so most of it its through GUI, but i removed the docker (its like docker rm) i did docker imager prune re"installed" docker (its like docker run) (as repository i have jc21/nginx-proxy-manager but on docker run it says its pulling from :latest) It didn't find available image so it pulled ALL layers from scratch. On WebGUI it still says 2.6.2. Is there some way to check from console? Maybe something in my config isn't compatible with 2.7.1 yet and this keeps it from upgrading? Just made a new container on different ports and new config mapping, still 2.6.2. I should probably start new issue , because i don't really have something that doesn't work, and its not relevant with this topic.
Author
Owner

@mgoeppl commented on GitHub (Nov 19, 2020):

2.7.1 works again for me ;)

<!-- gh-comment-id:730305268 --> @mgoeppl commented on GitHub (Nov 19, 2020): 2.7.1 works again for me ;)
Author
Owner

@jc21 commented on GitHub (Nov 22, 2020):

It's very strange that the UI would report a different version than the one running in your docker instance. Perhaps hit the API directly with a command like this:

curl http://yournpmhost:81/api

and see what the version number being reported is. If it's not the one you're expecting then you definitely have the wrong image running. If it's different to the UI, then maybe the browser is aggressively caching however there is code in the frontend to reload the page if the backend version is different than the one the UI was expecting.

If you are using NPM to proxy the NPM admin site itself (as most of us do) you will want to make sure the Cache Assets option is not selected.

<!-- gh-comment-id:731703486 --> @jc21 commented on GitHub (Nov 22, 2020): It's very strange that the UI would report a different version than the one running in your docker instance. Perhaps hit the API directly with a command like this: ```bash curl http://yournpmhost:81/api ``` and see what the version number being reported is. If it's not the one you're expecting then you definitely have the wrong image running. If it's different to the UI, then maybe the browser is aggressively caching however there is code in the frontend to reload the page if the backend version is different than the one the UI was expecting. If you are using NPM to proxy the NPM admin site itself (as most of us do) you will want to make sure the Cache Assets option is not selected.
Author
Owner

@stavros-k commented on GitHub (Nov 22, 2020):

It's very strange that the UI would report a different version than the one running in your docker instance. Perhaps hit the API directly with a command like this:

curl http://yournpmhost:81/api

and see what the version number being reported is. If it's not the one you're expecting then you definitely have the wrong image running. If it's different to the UI, then maybe the browser is aggressively caching however there is code in the frontend to reload the page if the backend version is different than the one the UI was expecting.

If you are using NPM to proxy the NPM admin site itself (as most of us do) you will want to make sure the Cache Assets option is not selected.

Ok, now i feel very stupid -.- I'm on Unraid and i'm using the templates there. which is a different image "jlesage/nginx-proxy-manager" i was seeing the "j" in the start and the "nginx-proxy-manager" after and i was like "yea this is ok"
I'll have either to wait until this image is updated, or try to convert your docker-compose to plain docker, or install portainer
Unraid doesn't support docker-compose natively,

<!-- gh-comment-id:731724323 --> @stavros-k commented on GitHub (Nov 22, 2020): > It's very strange that the UI would report a different version than the one running in your docker instance. Perhaps hit the API directly with a command like this: > > ```shell > curl http://yournpmhost:81/api > ``` > > and see what the version number being reported is. If it's not the one you're expecting then you definitely have the wrong image running. If it's different to the UI, then maybe the browser is aggressively caching however there is code in the frontend to reload the page if the backend version is different than the one the UI was expecting. > > If you are using NPM to proxy the NPM admin site itself (as most of us do) you will want to make sure the Cache Assets option is not selected. Ok, now i feel very **stupid** -.- I'm on Unraid and i'm using the templates there. which is a different image "jlesage/nginx-proxy-manager" i was seeing the "j" in the start and the "nginx-proxy-manager" after and i was like "yea this is ok" I'll have either to wait until this image is updated, or try to convert your docker-compose to plain docker, or install portainer Unraid doesn't support docker-compose natively,
Author
Owner

@LivingWithHippos commented on GitHub (Nov 25, 2020):

I had this issue after moving the variables to the compose file and removing config.json.
As jmeinicke said the solution was removing from the docker compose the line mounting the file

    volumes:
      - ./config.json:/app/config/production.json
<!-- gh-comment-id:733782994 --> @LivingWithHippos commented on GitHub (Nov 25, 2020): I had this issue after moving the variables to the compose file and removing `config.json`. As jmeinicke said the solution was removing from the docker compose the line mounting the file ```yaml volumes: - ./config.json:/app/config/production.json ```
Author
Owner

@mwip commented on GitHub (Dec 7, 2020):

I can confirm that is is still an issue with 2.7.1.

Rocking a Raspberry pi and just did a

sudo docker-compose down
sudo docker-compose pull
sudo docker-compose up -d

The proxy works just fine, but I cannot log in.

<!-- gh-comment-id:740196524 --> @mwip commented on GitHub (Dec 7, 2020): I can confirm that is is still an issue with 2.7.1. Rocking a Raspberry pi and just did a ``` sudo docker-compose down sudo docker-compose pull sudo docker-compose up -d ``` The proxy works just fine, but I cannot log in.
Author
Owner

@anselal commented on GitHub (Jan 4, 2021):

@mwip I get the same error on my RPi2.

<!-- gh-comment-id:753788807 --> @anselal commented on GitHub (Jan 4, 2021): @mwip I get the same error on my RPi2.
Author
Owner

@roswitina commented on GitHub (Jan 5, 2021):

I use a Helios64 with an arm64 board. The Docker image starts correctly, but the first time I login with the data admin@example.com/changeme I get the message "Bad Gateway".

The login page shows version 2.7.1.

I get the following error messages in the error-log:
2021/01/05 17:29:24 [error] 243#243: *336 connect() failed (111: Connection refused) while connecting to upstream, client: 127.0.0.1, server: nginxproxymanager, req

<!-- gh-comment-id:754779385 --> @roswitina commented on GitHub (Jan 5, 2021): I use a Helios64 with an arm64 board. The Docker image starts correctly, but the first time I login with the data admin@example.com/changeme I get the message "Bad Gateway". The login page shows version 2.7.1. I get the following error messages in the error-log: `2021/01/05 17:29:24 [error] 243#243: *336 connect() failed (111: Connection refused) while connecting to upstream, client: 127.0.0.1, server: nginxproxymanager, req`
Author
Owner

@dkupper76 commented on GitHub (Jan 18, 2021):

I just installed version 2.7.3, my first time using this, and I am having the same issues, login fails, and I get the "Bad Gateway" message.

Also, @jmeinicke mentioned the website: https://nginxproxymanager.com/setup/, but when I try to visit that website, I get the following error message from Firefox:

Secure Connection Failed

An error occurred during a connection to nginxproxymanager.com. SSL received a record that exceeded the maximum permissible length.

Error code: SSL_ERROR_RX_RECORD_TOO_LONG

Can you please help?

Thanks in advance.

<!-- gh-comment-id:762485695 --> @dkupper76 commented on GitHub (Jan 18, 2021): I just installed version 2.7.3, my first time using this, and I am having the same issues, login fails, and I get the "Bad Gateway" message. Also, @jmeinicke mentioned the website: https://nginxproxymanager.com/setup/, but when I try to visit that website, I get the following error message from Firefox: Secure Connection Failed An error occurred during a connection to nginxproxymanager.com. SSL received a record that exceeded the maximum permissible length. Error code: SSL_ERROR_RX_RECORD_TOO_LONG Can you please help? Thanks in advance.
Author
Owner

@dkupper76 commented on GitHub (Jan 18, 2021):

I forgot to add, I just looked at this in Portainer, and I see Container Unhealthy, and looking further I see the following error shown in Portainer:

"parse error: Invalid numeric literal at line 1, column 7 NOT OK"

I have not dug any deeper into that, but I thought I would mention it, because perhaps these issues are related.

Thanks in advance.

<!-- gh-comment-id:762486207 --> @dkupper76 commented on GitHub (Jan 18, 2021): I forgot to add, I just looked at this in Portainer, and I see Container Unhealthy, and looking further I see the following error shown in Portainer: "parse error: Invalid numeric literal at line 1, column 7 NOT OK" I have not dug any deeper into that, but I thought I would mention it, because perhaps these issues are related. Thanks in advance.
Author
Owner

@dkupper76 commented on GitHub (Jan 20, 2021):

I tried downgrading to version 2.7.1, which somebody said was working, but I still get the same issue, and it actually shows up as version 2.7.2. The Website : https://nginxproxymanager.com/setup/, is still having certificate errors, and I can't visit it, so I opened a new issue on the website cert errors

<!-- gh-comment-id:763615252 --> @dkupper76 commented on GitHub (Jan 20, 2021): I tried downgrading to version 2.7.1, which somebody said was working, but I still get the same issue, and it actually shows up as version 2.7.2. The Website : https://nginxproxymanager.com/setup/, is still having certificate errors, and I can't visit it, so I opened a new issue on the website cert errors
Author
Owner

@dkupper76 commented on GitHub (Jan 28, 2021):

I solved the problem by deleting the folders on the Host Machine running Docker, deleting the docker compose file and starting from scratch, not sure what caused my problem, but it appears to be working now.

<!-- gh-comment-id:769082709 --> @dkupper76 commented on GitHub (Jan 28, 2021): I solved the problem by deleting the folders on the Host Machine running Docker, deleting the docker compose file and starting from scratch, not sure what caused my problem, but it appears to be working now.
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
starred/nginx-proxy-manager-NginxProxyManager#615
No description provided.