FAQ

Why is Traefik Answering XXX HTTP Response Status Code?

Traefik is a dynamic reverse proxy, and while the documentation often demonstrates configuration options through file examples, the core feature of Traefik is its dynamic configurability, directly reacting to changes from providers over time.

Notably, a part of the configuration is static, and can be provided by a file on startup, whereas various providers, such as the file provider, contribute dynamically all along the traefik instance lifetime to its dynamic configuration changes.

In addition, the configuration englobes concepts such as the EntryPoint which can be seen as a listener on the Transport Layer (TCP), as apposed to the Router which is more about the Presentation (TLS) and Application layers (HTTP). And there can be as many routers as one wishes for a given EntryPoint.

In other words, for a given Entrypoint, at any given time the traffic seen is not bound to be just about one protocol. It could be HTTP, or otherwise. Over TLS, or not. Not to mention that dynamic configuration changes potentially make that kind of traffic vary over time.

Therefore, in this dynamic context, the static configuration of an entryPoint does not give any hint whatsoever about how the traffic going through that entryPoint is going to be routed. Or whether it’s even going to be routed at all, that is whether there is a Router matching the kind of traffic going through it.

404 Not found

Traefik returns a 404 response code in the following situations:

• A request reaching an EntryPoint that has no Routers
• An HTTP request reaching an EntryPoint that has no HTTP Router
• An HTTPS request reaching an EntryPoint that has no HTTPS Router
• A request reaching an EntryPoint that has HTTP/HTTPS Routers that cannot be matched

From Traefik’s point of view, every time a request cannot be matched with a router the correct response code is a 404 Not found.

In this situation, the response code is not a 503 Service Unavailable because Traefik is not able to confirm that the lack of a matching router for a request is only temporary. Traefik’s routing configuration is dynamic and aggregated from different providers, hence it’s not possible to assume at any moment that a specific route should be handled or not.

This behavior is consistent with rfc7231

The server is currently unable to handle the request due to a
temporary overloading or maintenance of the server. The implication
is that this is a temporary condition which will be alleviated after
some delay. If known, the length of the delay MAY be indicated in a
Retry-After header. If no Retry-After is given, the client SHOULD
handle the response as it would for a 500 response.
    Note: The existence of the 503 status code does not imply that a
    server must use it when becoming overloaded. Some servers may wish
    to simply refuse the connection.

Extract from rfc7231#section-6.6.4.

502 Bad Gateway

Traefik returns a 502 response code when an error happens while contacting the upstream service.

503 Service Unavailable

Traefik returns a 503 response code when a Router has been matched, but there are no servers ready to handle the request.

This situation is encountered when a service has been explicitly configured without servers, or when a service has healthcheck enabled and all servers are unhealthy.

XXX Instead of 404

Sometimes, the 404 response code doesn’t play well with other parties or services (such as CDNs).

In these situations, you may want Traefik to always reply with a 503 response code, instead of a 404 response code.

To achieve this behavior, a catchall router, with the lowest possible priority and routing to a service without servers, can handle all the requests when no other router has been matched.

The example below is a file provider only version (yaml) of what this configuration could look like:

Static configuration

# traefik.yml
entryPoints:
  web:
    address: :80
providers:
  file:
    filename: dynamic.yaml

Dynamic configuration

# dynamic.yaml
http:
  routers:
    catchall:
      # attached only to web entryPoint
      entryPoints:
        - "web"
      # catchall rule
      rule: "PathPrefix(`/`)"
      service: unavailable
      # lowest possible priority
      # evaluated when no other router is matched
      priority: 1
  services:
    # Service that will always answer a 503 Service Unavailable response
    unavailable:
      loadBalancer:
        servers: {}

Dedicated service

If there is a need for a response code other than a 503 and/or a custom message, the principle of the above example above (a catchall router) still stands, but the unavailable service should be adapted to fit such a need.

Why Is My TLS Certificate Not Reloaded When Its Contents Change?

With the file provider, a configuration update is only triggered when one of the watched configuration files is modified.

Which is why, when a certificate is defined by path, and the actual contents of this certificate change, a configuration update is not triggered.

To take into account the new certificate contents, the update of the dynamic configuration must be forced. One way to achieve that, is to trigger a file notification, for example, by using the touch command on the configuration file.

What Are the Forwarded Headers When Proxying HTTP Requests?

By default, the following headers are automatically added when proxying requests:

For more details, please check out the forwarded header documentation.

How Traefik is Storing and Serving TLS Certificates?

Storing TLS Certificates

TLS certificates are either provided directly by the dynamic configuration from providers, or by ACME resolvers, which act themselves as providers internally.

For each TLS certificate, Traefik produces an identifier used as a key to store it. This identifier is constructed as the alphabetically ordered concatenation of the SANs DNSNames and IPAddresses of the TLScertificate.

Examples:

The identifier is used to store TLS certificates in order to be later used to handle TLS connections. This operation happens each time there are configuration changes.

If multiple TLS certificates are provided with the same SANs definition (same identifier), only the one processed first is kept. Because the dynamic configuration is aggregated from all providers, when processing it to gather TLS certificates, there is no guarantee of the order in which they would be processed. This means that along with configurations applied, it is possible that the TLS certificate retained for a given identifier differs.

Serving TLS Certificates

For each incoming connection, Traefik is serving the “best” matching TLS certificate for the provided server name.

The TLS certificate selection process narrows down the list of TLS certificates matching the server name, and then selects the last TLS certificate in this list after having ordered it by the identifier alphabetically.

Examples:

Caching TLS Certificates

While Traefik is serving the best matching TLS certificate for each incoming connection, the selection process cost for each incoming connection is avoided thanks to a cache mechanism.

Once a TLS certificate has been selected as the “best” TLS certificate for a server name, it is cached for an hour, avoiding the selection process for further connections.

Nonetheless, when a new configuration is applied, the cache is reset.

What does the “field not found” error mean?

error: field not found, node: -badField-

The “field not found” error occurs, when an unknown property is encountered in the dynamic or static configuration.

One way to check whether a configuration file is well-formed, is to validate it with:

• JSON Schema of the static configuration
• JSON Schema of the dynamic configuration

Why are some resources (routers, middlewares, services…) not created/applied?

As a common tip, if a resource is dropped/not created by Traefik after the dynamic configuration was evaluated, one should look for an error in the logs.

If found, the error confirms that something went wrong while creating the resource, and the message should help in figuring out the mistake(s) in the configuration, and how to fix it.

When using the file provider, one way to check if the dynamic configuration is well-formed is to validate it with the JSON Schema of the dynamic configuration.

Why does Let’s Encrypt wildcard certificate renewal/generation with DNS challenge fail?

If you’re trying to renew wildcard certificates, with DNS challenge, and you’re getting errors such as:

msg="Error renewing certificate from LE: {example.com [*.example.com]}"
providerName=letsencrypt.acme error="error: one or more domains had a problem:
[example.com] acme: error presenting token: gandiv5: unexpected authZone example.com. for fqdn example.com."

then it could be due to CNAME support.

In which case, you should make sure your infrastructure is properly set up for a DNS challenge that does not rely on CNAME, and you should try disabling CNAME support with:

LEGO_DISABLE_CNAME_SUPPORT=true