This document contains common problems and their solutions.
If you came across an issue where the solution was not immediately obvious, consider adding it to this list to help other users.
- Refused to Connect in Web Content View
- 404 On Static Hosting
- Yarn Build or Run Error
- Auth Validation Error: "should be object"
- Keycloak Redirect Error
- Docker Directory Error
- Config Not Updating
- Config Still not Updating
- Styles and Assets not Updating
- DockerHub toomanyrequests
- Config Validation Errors
- Node Sass unsupported environment
- Cannot find module './_baseValues'
- Ngrok Invalid Host Headers
- Warnings in the Console during deploy
- Docker Login Fails on Ubuntu
- Status Checks Failing
- Diagnosing Widget Errors
- Fixing Widget CORS Errors
- Keycloak Redirect Error
- How-To Open Browser Console
- Git Contributions not Displaying
This is not an issue with Dashy, but instead caused by the target app preventing direct access through embedded elements. It can be fixed by setting the X-Frame-Options
HTTP header set to ALLOW [path to Dashy]
or SAMEORIGIN
, as defined in RFC-7034. These settings are usually set in the config file for the web server that's hosting the target application, here are some examples of how to enable cross-origin access with common web servers:
In NGINX, you can use the add_header
module within the app block.
server {
...
add_header X-Frame-Options SAMEORIGIN always;
}
Then reload with service nginx reload
In Caddy, you can use the header
directive.
header {
X-Frame-Options SAMEORIGIN
}
In Apache, you can use the mod_headers
module to set the X-Frame-Options
in your config file. This file is usually located somewhere like `/etc/apache2/httpd.conf
Header set X-Frame-Options: "ALLOW-FROM http://[dashy-location]/"
If you're seeing Dashy's 404 page on initial load/ refresh, and then the main app when you go back to Home, then this is likely caused by the Vue router, and if so can be fixed in one of two ways.
The first solution is to switch the routing mode, from HTML5 history
mode to hash
mode, by setting appConfig.routingMode
to hash
.
If this works, but you wish to continue using HTML5 history mode, then a bit of extra server configuration is required. This is explained in more detaail in the Vue Docs. Once completed, you can then use routingMode: history
again, for neater URLs.
For more info, see Issue #1
First of all, check that you've got yarn installed correctly - see the yarn installation docs for more info.
If you're getting an error about scenarios, then you've likely installed the wrong yarn... (you're not the only one!). You can fix it by uninstalling, adding the correct repo, and reinstalling, for example, in Debian:
sudo apt remove yarn
curl -sS https://dl.yarnpkg.com/debian/pubkey.gpg | sudo apt-key add -
echo "deb https://dl.yarnpkg.com/debian/ stable main" | sudo tee /etc/apt/sources.list.d/yarn.list
sudo apt update && sudo apt install yarn
Alternatively, as a workaround, you have several options:
- Try using NPM instead: So clone, cd, then run
npm install
,npm run build
andnpm start
- Try using Docker instead, and all of the system setup and dependencies will already be taken care of. So from within the directory, just run
docker build -t lissy93/dashy .
to build, and then use docker start to run the project, e.g:docker run -it -p 8080:80 lissy93/dashy
(see the deploying docs for more info)
In V 1.6.5 an update was made that in the future will become a breaking change. You will need to update you config to reflect this before V 2.0.0 is released. In the meantime, your previous config will continue to function normally, but you will see a validation warning. The change means that the structure of the appConfig.auth
object is now an object, which has a users
property.
For more info, see this announcement.
You can fix this by replacing:
auth:
- user: xxx
hash: xxx
with
auth:
users:
- user: xxx
hash: xxx
Check the browser's console output, if you've not set any headers, you will likely see a CORS error here, which would be the source of the issue.
You need to allow Dashy to make requests to Keycloak, and Keycloak to redirect to Dashy. The way you do this depends on how you're hosting these applications / which proxy you are using, and examples can be found in the Management Docs.
For example, add the access control header to Keycloak, like:
Access-Control-Allow-Origin [URL-of Dashy]
Note that for requests that transport sensitive info like credentials, setting the accept header to a wildcard (*
) is not allowed - see MDN Docs, so you will need to specify the actual URL.
You should also ensure that Keycloak is correctly configured, with a user, realm and application, and be sure that you have set a valid redirect URL in Keycloak (screenshot).
For more details on how to set headers, see the Example Headers in the management docs, or reference the documentation for your proxy.
See also: #479, #409, #507, #491, #341, #520
Error response from daemon: OCI runtime create failed: container_linux.go:380:
starting container process caused: process_linux.go:545: container init caused:
rootfs_linux.go:76: mounting "/home/ubuntu/my-conf.yml" to rootfs at
"/app/public/conf.yml" caused: mount through procfd: not a directory:
unknown: Are you trying to mount a directory onto a file (or vice-versa)?
Check if the specified host path exists and is the expected type.
If you get an error similar to the one above, you are mounting a directory to the config file's location, when a plain file is expected. Create a YAML file, (touch my-conf.yml
), populate it with a sample config, then pass it as a volume: -v ./my-local-conf.yml:/app/public/conf.yml
Dashy has the option to save settings and config locally, in browser storage. Anything here will take precedence over whatever is in your config file, sometimes with unintended consequences. If you've updated the config file manually, and are not seeing changes reflected in the UI, then try visiting the site in Incognito mode. If that works, then the solution is just to clear local storage. This can be done from the config menu, under "Clear Local Settings".
Sometimes your text editor updates files inode, meaning changes will not be picked up by the Docker container. This article explains things further.
If you find that your styles and other visual assets work when visiting ip:port
by not dashy.domain.com
, then this is usually caused by caching. In your browser, do a hard-refresh (Ctrl + F5). If you use Cloudflare, then you can clear the cache through the management console, or set the cache level to Bypass for certain files, under the Rules tab.
This situation relates to error messages similar to one of the following, returned when pulling, updating or running the Docker container from Docker Hub.
Continuing execution. Pulling image lissy93/dashy:release-1.6.0
error pulling image configuration: toomanyrequests
or
You have reached your pull rate limit. You may increase the limit by authenticating and upgrading: https://www.docker.com/increase-rate-limit
When DockerHub returns one of these errors, or a 429
status, that means you've hit your rate limit. This was introduced last year, and prevents unauthenticated or free users from running docker pull more than 100 times per 6 hours.
You can check your rate limit status by looking for the ratelimit-remaining
header in any DockerHub responses.
- Dashy is also availible through GHCR, which at present does not have any hard limits. Just use
docker pull ghcr.io/lissy93/dashy:latest
to fetch the image - You can also build the image from source, by cloning the repo, and running
docker build -t dashy .
or use the pre-made docker compose
- Logging in to DockerHub will increase your rate limit from 100 requests to 200 requests per 6 hour period
- Upgrading to a Pro for $5/month will increase your image requests to 5,000 per day, and any plans above have no rate limits
- Since rate limits are counted based on your IP address, proxying your requests, or using a VPN may work
The configuration file is validated against Dashy's Schema using AJV.
First, check that your syntax is valid, using YAML Validator or JSON Validator. If the issue persists, then take a look at the schema, and verify that the field you are trying to add/ modify matches the required format. You can also use this tool to validate your JSON config against the schema, or run yarn validate-config
.
If you're trying to use a recently released feature, and are getting a warning, this is likely because you've not yet updated the the current latest version of Dashy.
If the issue still persists, you should raise an issue.
Caused by node-sass's binaries being built for a for a different architecture
To fix this, just run: yarn rebuild node-sass
Clearing the cache should fix this: yarn cache clean
If the issue persists, remove (rm -rf node_modules\ yarn.lock
) and reinstall (yarn
) node_modules
Just add the -host-header flag, e.g. ngrok http 8080 -host-header="localhost:8080"
Please acknowledge the difference between errors and warnings before raising an issue about messages in the console. It's not unusual to see warnings about a new version of a certain package being available, an asset bundle bing oversized or a service worker not yet having a cache. These shouldn't have any impact on the running application, so please don't raise issues about these unless it directly relates to a bug or issue you're experiencing. Errors on the other hand should not appear in the console, and they are worth looking into further.
Run sudo apt install gnupg2 pass && gpg2 -k
If you're using status checks, and despite a given service being online, the check is displaying an error, there are a couple of things you can look at:
If your service requires requests to include any authorization in the headers, then use the statusCheckHeaders
property, as described in the docs.
If you are still having issues, it may be because your target application is blocking requests from Dashy's IP. This is a CORS error, and can be fixed by setting the headers on your target app, to include:
Access-Control-Allow-Origin: https://location-of-dashy/
Vary: Origin
If the URL you are checking has an unsigned certificate, or is not using HTTPS, then you may need to disable the rejection of insecure requests. This can be done by setting statusCheckAllowInsecure
to true for a given item.
If your service is online, but responds with a status code that is not in the 2xx range, then you can use statusCheckAcceptCodes
to set an accepted status code.
If you get an error, like Service Unavailable: Server resulted in a fatal error
, even when it's definitely online, this is most likely caused by missing the protocol. Don't forget to include https://
(or whatever protocol) before the URL, and ensure that if needed, you've specified the port.
Running Dashy in HOST network mode, instead of BRIDGE will allow status check access to other services in HOST mode. For more info, see #445.
If you have firewall rules configured, then ensure that they don't prevent Dashy from making requests to the other services you are trying to access.
Currently, the status check needs a page to be rendered, so if this URL in your browser does not return anything, then status checks will not work. This may be modified in the future, but in the meantime, a fix would be to make your own status service, which just checks if your app responds with whatever code you'd like, and then return a 200 plus renders an arbitrary message. Then just point statusCheckUrl
to your custom page.
For further troubleshooting, use an application like Postman to diagnose the issue. Set the parameter to GET
, and then make a call to: https://[url-of-dashy]/status-check/?&url=[service-url]
. Where the service URL must have first been encoded (e.g. with encodeURIComponent()
or urlencoder.io)
If you're serving Dashy though a CDN, instead of using the Node server or Docker image, then the Node endpoint that makes requests will not be available to you, and all requests will fail. A workaround for this may be implemented in the future, but in the meantime, your only option is to use the Docker or Node deployment method.
If an error occurs when fetching or rendering results, you will see a short message in the UI. If that message doesn't addequatley explain the problem, then you can open the browser console to see more details.
Before proceeding, ensure that if the widget requires auth your API is correct, and for custom widgets, double check that the URL and protocol is correct.
If the console message mentions to corss-origin blocking, then this is a CORS error, see: Fixing Widget CORS Errors
If you're able to, you can find more information about why the request may be failing in the Dev Tools under the Network tab, and you can ensure your endpoint is correct and working using a tool like Postman.
The most common widget issue is a CORS error. This is a browser security mechanism which prevents the client-side app (Dashy) from from accessing resources on a remote origin, without that server's explicit permission (e.g. with headers like Access-Control-Allow-Origin). See the MDN Docs for more info: Cross-Origin Resource Sharing.
There are several ways to fix a CORS error:
You will get a CORS error if you try and access a http service from a https source. So ensure that the URL you are requesting has the right protocol, and is correctly formatted.
If you have control over the destination (e.g. for a self-hosted service), then you can simply apply the correct headers.
Add the Access-Control-Allow-Origin
header, with the value of either *
to allow requests from anywhere, or more securely, the host of where Dashy is served from. For example:
Access-Control-Allow-Origin: https://url-of-dashy.local
or
Access-Control-Allow-Origin: *
For more info on how to set headers, see: Setting Headers in the management docs
You can route requests through Dashy's built-in CORS proxy. Instructions and more details can be found here. If you don't have control over the target origin, and you are running Dashy either through Docker, with the Node server or on Netlify, then this solution will work for you.
Just add the useProxy: true
option to the failing widget.
For testing purposes, you can use an addon, which will disable the CORS checks. You can get the Allow-CORS extension for Chrome or Firefox, more details here
Firstly, ensure that in your Keycloak instance you have populated the Valid Redirect URIs field (screenshot) with the URL to your Dashy instance.
You may need to specify CORS headers on your Keycloak instance, to allow requests coming from Dashy, e.g:
Access-Control-Allow-Origin: https://dashy.example.com
If you're running in Kubernetes, you will need to enable CORS ingress rules, see docs, e.g:
nginx.ingress.kubernetes.io/cors-allow-origin: "https://dashy.example.com"
nginx.ingress.kubernetes.io/enable-cors: "true"
When raising a bug, one crucial piece of info needed is the browser's console output. This will help the developer diagnose and fix the issue.
If you've been asked for this info, but are unsure where to find it, then it is under the "Console" tab, in the browsers developer tools, which can be opened with F12. You can right-click the console, and select Save As to download the log.
To open dev tools, and jump straight to the console:
- Win / Linux: Ctrl + Shift + J
- MacOS: Cmd + Option + J
For more detailed walk through, see this article.
If you've contributed to Dashy (or any other project), but your contributions are not showing up on your GH profile, or in Dashy's Credits Page, then this is likely a git config issue.
These statistics are generated using the username / email associated with commits. This info needs to be setup on your local machine using git config
.
Run the following commands (replacing name + email with your info):
git config --global user.name "John Doe"
git config --global user.email [email protected]
For more info, see Git First Time Setup Docs.
Note that only contributions to the master / main branch or a project are counted