Impact

Contour's Cookie Rewriting feature is vulnerable to Lua code injection. An attacker with RBAC permissions to create or modify HTTPProxy resources can craft a malicious value in the following fields that results in arbitrary code execution in the Envoy proxy:

- spec.routes[].cookieRewritePolicies[].pathRewrite.value - spec.routes[].services[].cookieRewritePolicies[].pathRewrite.value

The cookie rewriting feature is internally implemented using Envoy's HTTP Lua filter. User-controlled values are interpolated into Lua source code using Go text/template without sufficient sanitization.

The injected code only executes when processing traffic on the attacker's own route, which they already control. However, since Envoy runs as shared infrastructure, the injected code can also:

- Read Envoy's xDS client credentials from the filesystem, which could be used to read all Contour xDS configuration, including TLS certificates and private keys of other tenants. - Cause denial of service for other tenants sharing the Envoy instance.

Other use cases of Lua filter are not vulnerable.

Patches

The fix is available in Contour v1.33.4, v1.32.5, and v1.31.6.

- v1.33.4: User-provided values are no longer interpolated into Lua code. Use of text/template is removed. Requires Envoy 1.35.0 or later. - v1.32.5, v1.31.6: User-provided values are escaped before interpolation into Lua code.

Workarounds

There are no workarounds. Users should upgrade to a patched version.

Insecure permissions in contour v1.28.3 allows attackers to access sensitive data and escalate privileges by obtaining the service account's token.

HTTP/2 Rapid reset attack The HTTP/2 protocol allows clients to indicate to the server that a previous stream should be canceled by sending a RSTSTREAM frame. The protocol does not require the client and server to coordinate the cancellation in any way, the client may do it unilaterally. The client may also assume that the cancellation will take effect immediately when the server receives the RSTSTREAM frame, before any other data from that TCP connection is processed.

Abuse of this feature is called a Rapid Reset attack because it relies on the ability for an endpoint to send a RSTSTREAM frame immediately after sending a request frame, which makes the other endpoint start working and then rapidly resets the request. The request is canceled, but leaves the HTTP/2 connection open.

The HTTP/2 Rapid Reset attack built on this capability is simple: The client opens a large number of streams at once as in the standard HTTP/2 attack, but rather than waiting for a response to each request stream from the server or proxy, the client cancels each request immediately.

The ability to reset streams immediately allows each connection to have an indefinite number of requests in flight. By explicitly canceling the requests, the attacker never exceeds the limit on the number of concurrent open streams. The number of in-flight requests is no longer dependent on the round-trip time (RTT), but only on the available network bandwidth.

In a typical HTTP/2 server implementation, the server will still have to do significant amounts of work for canceled requests, such as allocating new stream data structures, parsing the query and doing header decompression, and mapping the URL to a resource. For reverse proxy implementations, the request may be proxied to the backend server before the RSTSTREAM frame is processed. The client on the other hand paid almost no costs for sending the requests. This creates an exploitable cost asymmetry between the server and the client.

Multiple software artifacts implementing HTTP/2 are affected. This advisory was originally ingested from the swift-nio-http2 repo advisory and their original conent follows.

swift-nio-http2 specific advisory swift-nio-http2 is vulnerable to a denial-of-service vulnerability in which a malicious client can create and then reset a large number of HTTP/2 streams in a short period of time. This causes swift-nio-http2 to commit to a large amount of expensive work which it then throws away, including creating entirely new Channels to serve the traffic. This can easily overwhelm an EventLoop and prevent it from making forward progress.

swift-nio-http2 1.28 contains a remediation for this issue that applies reset counter using a sliding window. This constrains the number of stream resets that may occur in a given window of time. Clients violating this limit will have their connections torn down. This allows clients to continue to cancel streams for legitimate reasons, while constraining malicious actors.

Contour is a Kubernetes ingress controller using Envoy proxy. In Contour before version 1.17.1 a specially crafted ExternalName type Service may be used to access Envoy's admin interface, which Contour normally prevents from access outside the Envoy container. This can be used to shut down Envoy remotely (a denial of service), or to expose the existence of any Secret that Envoy is using for its configuration, including most notably TLS Keypairs. However, it cannot be used to get the content of those secrets. Since this attack allows access to the administration interface, a variety of administration options are available, such as shutting down the Envoy or draining traffic. In general, the Envoy admin interface cannot easily be used for making changes to the cluster, in-flight requests, or backend services, but it could be used to shut down or drain Envoy, change traffic routing, or to retrieve secret metadata, as mentioned above. The issue will be addressed in Contour v1.18.0 and a cherry-picked patch release, v1.17.1, has been released to cover users who cannot upgrade at this time. For more details refer to the linked GitHub Security Advisory.

In Contour ( Ingress controller for Kubernetes) before version 1.7.0, a bad actor can shut down all instances of Envoy, essentially killing the entire ingress data plane. GET requests to /shutdown on port 8090 of the Envoy pod initiate Envoy's shutdown procedure. The shutdown procedure includes flipping the readiness endpoint to false, which removes Envoy from the routing pool. When running Envoy (For example on the host network, pod spec hostNetwork=true), the shutdown manager's endpoint is accessible to anyone on the network that can reach the Kubernetes node that's running Envoy. There is no authentication in place that prevents a rogue actor on the network from shutting down Envoy via the shutdown manager endpoint. Successful exploitation of this issue will lead to bad actors shutting down all instances of Envoy, essentially killing the entire ingress data plane. This is fixed in version 1.7.0.