Vulnerabilities > CVE-2024-43414 - Uncontrolled Recursion vulnerability in Apollographql products
Summary
Apollo Federation is an architecture for declaratively composing APIs into a unified graph. Each team can own their slice of the graph independently, empowering them to deliver autonomously and incrementally. Instances of @apollo/query-planner >=2.0.0 and <2.8.5 are impacted by a denial-of-service vulnerability. @apollo/gateway versions >=2.0.0 and < 2.8.5 and Apollo Router <1.52.1 are also impacted through their use of @apollo/query-panner. If @apollo/query-planner is asked to plan a sufficiently complex query, it may loop infinitely and never complete. This results in unbounded memory consumption and either a crash or out-of-memory (OOM) termination. This issue can be triggered if you have at least one non-@key field that can be resolved by multiple subgraphs. To identify these shared fields, the schema for each subgraph must be reviewed. The mechanism to identify shared fields varies based on the version of Federation your subgraphs are using. You can check if your subgraphs are using Federation 1 or Federation 2 by reviewing their schemas. Federation 2 subgraph schemas will contain a @link directive referencing the version of Federation being used while Federation 1 subgraphs will not. For example, in a Federation 2 subgraph, you will find a line like @link(url: "https://specs.apollo.dev/federation/v2.0"). If a similar @link directive is not present in your subgraph schema, it is using Federation 1. Note that a supergraph can contain a mix of Federation 1 and Federation 2 subgraphs. This issue results from the Apollo query planner attempting to use a Number exceeding Javascript’s Number.MAX_VALUE in some cases. In Javascript, Number.MAX_VALUE is (2^1024 - 2^971). When the query planner receives an inbound graphql request, it breaks the query into pieces and for each piece, generates a list of potential execution steps to solve the piece. These candidates represent the steps that the query planner will take to satisfy the pieces of the larger query. As part of normal operations, the query planner requires and calculates the number of possible query plans for the total query. That is, it needs the product of the number of query plan candidates for each piece of the query. Under normal circumstances, after generating all query plan candidates and calculating the number of all permutations, the query planner moves on to stack rank candidates and prune less-than-optimal options. In particularly complex queries, especially those where fields can be solved through multiple subgraphs, this can cause the number of all query plan permutations to balloon. In worst-case scenarios, this can end up being a number larger than Number.MAX_VALUE. In Javascript, if Number.MAX_VALUE is exceeded, Javascript represents the value as “infinity”. If the count of candidates is evaluated as infinity, the component of the query planner responsible for pruning less-than-optimal query plans does not actually prune candidates, causing the query planner to evaluate many orders of magnitude more query plan candidates than necessary. This issue has been addressed in @apollo/query-planner v2.8.5, @apollo/gateway v2.8.5, and Apollo Router v1.52.1. Users are advised to upgrade. This issue can be avoided by ensuring there are no fields resolvable from multiple subgraphs. If all subgraphs are using Federation 2, you can confirm that you are not impacted by ensuring that none of your subgraph schemas use the @shareable directive. If you are using Federation 1 subgraphs, you will need to validate that there are no fields resolvable by multiple subgraphs.
Vulnerable Configurations
Common Weakness Enumeration (CWE)
Common Attack Pattern Enumeration and Classification (CAPEC)
- XML Nested Payloads Applications often need to transform data in and out of the XML format by using an XML parser. It may be possible for an attacker to inject data that may have an adverse effect on the XML parser when it is being processed. By nesting XML data and causing this data to be continuously self-referential, an attacker can cause the XML parser to consume more resources while processing, causing excessive memory consumption and CPU utilization. An attacker's goal is to leverage parser failure to his or her advantage. In most cases this type of an attack will result in a denial of service due to an application becoming unstable, freezing, or crash. However it may be possible to cause a crash resulting in arbitrary code execution, leading to a jump from the data plane to the control plane [R.230.1].
- XML Oversized Payloads Applications often need to transform data in and out of the XML format by using an XML parser. It may be possible for an attacker to inject data that may have an adverse effect on the XML parser when it is being processed. By supplying oversized payloads in input vectors that will be processed by the XML parser, an attacker can cause the XML parser to consume more resources while processing, causing excessive memory consumption and CPU utilization, and potentially cause execution of arbitrary code. An attacker's goal is to leverage parser failure to his or her advantage. In many cases this type of an attack will result in a denial of service due to an application becoming unstable, freezing, or crash. However it is possible to cause a crash resulting in arbitrary code execution, leading to a jump from the data plane to the control plane [R.231.1].
- Violating Implicit Assumptions Regarding XML Content (aka XML Denial of Service (XDoS)) XML Denial of Service (XDoS) can be applied to any technology that utilizes XML data. This is, of course, most distributed systems technology including Java, .Net, databases, and so on. XDoS is most closely associated with web services, SOAP, and Rest, because remote service requesters can post malicious XML payloads to the service provider designed to exhaust the service provider's memory, CPU, and/or disk space. The main weakness in XDoS is that the service provider generally must inspect, parse, and validate the XML messages to determine routing, workflow, security considerations, and so on. It is exactly these inspection, parsing, and validation routines that XDoS targets. There are three primary attack vectors that XDoS can navigate Target CPU through recursion: attacker creates a recursive payload and sends to service provider Target memory through jumbo payloads: service provider uses DOM to parse XML. DOM creates in memory representation of XML document, but when document is very large (for example, north of 1 Gb) service provider host may exhaust memory trying to build memory objects. XML Ping of death: attack service provider with numerous small files that clog the system. All of the above attacks exploit the loosely coupled nature of web services, where the service provider has little to no control over the service requester and any messages the service requester sends.
- XML Parser Attack Applications often need to transform data in and out of the XML format by using an XML parser. It may be possible for an attacker to inject data that may have an adverse effect on the XML parser when it is being processed. These adverse effects may include the parser crashing, consuming too much of a resource, executing too slowly, executing code supplied by an attacker, allowing usage of unintended system functionality, etc. An attacker's goal is to leverage parser failure to his or her advantage. In some cases it may be possible to jump from the data plane to the control plane via bad data being passed to an XML parser. [R.99.1]