CVE-2024-32971

Apollo Router is a configurable, graph router written in Rust to run a federated supergraph that uses Apollo Federation 2. The affected versions of Apollo Router contain a bug that in limited circumstances, could lead to unexpected operations being executed which can result in unintended data or effects. This only affects Router instances configured to use distributed query plan caching. The root cause of this defect is a bug in Apollo Router’s cache retrieval logic: When this defect is present and distributed query planning caching is enabled, asking the Router to execute an operation (whether it is a query, a mutation, or a subscription) may result in an unexpected variation of that operation being executed or the generation of unexpected errors. The issue stems from inadvertently executing a modified version of a previously executed operation, whose query plan is stored in the underlying cache (specifically, Redis). Depending on the type of the operation, the result may vary. For a query, results may be fetched that don’t match what was requested (e.g., rather than running `fetchUsers(type: ENTERPRISE)` the Router may run `fetchUsers(type: TRIAL)`. For a mutation, this may result in incorrect mutations being sent to underlying subgraph servers (e.g., rather than sending `deleteUser(id: 10)` to a subgraph, the Router may run `deleteUser(id: 12)`. Users who are using distributed query plan caching, are advised to either upgrade to version 1.45.1 or above or downgrade to version 1.43.2 of the Apollo Router. Apollo Router versions 1.44.0 or 1.45.0 are not recommended for use and have been withdrawn. Users unable to upgrade can disable distributed query plan caching to mitigate this issue.

Apollo Router is a configurable, graph router written in Rust to run a federated supergraph that uses Apollo Federation 2. The affected versions of Apollo Router contain a bug that in limited circumstances, could lead to unexpected operations being executed which can result in unintended data or effects. This only affects Router instances configured to use distributed query plan caching. The root cause of this defect is a bug in Apollo Router’s cache retrieval logic: When this defect is present and distributed query planning caching is enabled, asking the Router to execute an operation (whether it is a query, a mutation, or a subscription) may result in an unexpected variation of that operation being executed or the generation of unexpected errors. The issue stems from inadvertently executing a modified version of a previously executed operation, whose query plan is stored in the underlying cache (specifically, Redis). Depending on the type of the operation, the result may vary. For a query, results may be fetched that don’t match what was requested (e.g., rather than running `fetchUsers(type: ENTERPRISE)` the Router may run `fetchUsers(type: TRIAL)`. For a mutation, this may result in incorrect mutations being sent to underlying subgraph servers (e.g., rather than sending `deleteUser(id: 10)` to a subgraph, the Router may run `deleteUser(id: 12)`. Users who are using distributed query plan caching, are advised to either upgrade to version 1.45.1 or above or downgrade to version 1.43.2 of the Apollo Router. Apollo Router versions 1.44.0 or 1.45.0 are not recommended for use and have been withdrawn. Users unable to upgrade can disable distributed query plan caching to mitigate this issue.

### Impact Any instance of Apollo Router 1.44.0 or 1.45.0 that is using Distributed Query Plan Caching is impacted. These versions were released on 2024-04-12 and 2024-04-22 respectively. The affected versions of Apollo Router contain a bug that could lead to unexpected operations being executed, which can result in unintended data or effects. This only affects Router instances configured to use distributed query plan caching. Router versions other than the ones listed above, and all Router deployments that are not using distributed query plan caching, are unaffected by this defect. If you are using the affected versions, you can check your router’s configuration YAML to verify if you are impacted: ```yaml supergraph: query_planning: cache: # Look for this config below redis: urls: ["redis://..."] ``` A full reference on the[ Distributed Query Plan Caching feature is available here.](https://www.apollographql.com/docs/router/configuration/distributed-caching/#distributed-query-plan-caching) ### Impact detail The root cause of this defect is a bug in Apollo Router’s cache retrieval logic: When this defect is present and distributed query planning caching is enabled, asking the Router to execute an operation (whether it is a query, a mutation, or a subscription) may result in an unexpected variation of that operation being executed or the generation of unexpected errors. The issue stems from inadvertently executing a modified version of a previously executed operation, whose query plan is stored in the underlying cache (specifically, Redis). Depending on the type of the operation, the result may vary. For a query, results may be fetched that don’t match what was requested (e.g., rather than running `fetchUsers(type: ENTERPRISE)` the Router may run `fetchUsers(type: TRIAL)`. For a mutation, this may result in incorrect mutations being sent to underlying subgraph servers (e.g., rather than sending `deleteUser(id: 10)` to a subgraph, the Router may run `deleteUser(id: 12)`. ### Patches Apollo Router 1.45.1 If you are using distributed query plan caching, please either upgrade to version 1.45.1 or above or downgrade to version 1.43.2 of the Apollo Router. We do not recommend Apollo Router versions 1.44.0 or 1.45.0 for use and have withdrawn these releases. If you use impacted versions in production, we recommend that you migrate away immediately by redeploying to an unaffected Router version. For non-production use cases, we recommend you migrate at your earliest convenience. ### Workarounds If you cannot upgrade or downgrade, you can disable distributed query plan caching by removing the `supergraph.query_planning.cache.redis.urls` configuration. Please note that when distributed query plan caching is disabled, each Router instance will maintain its own in-memory query plan cache. This may increase resource utilization for each Router instance and could increase cold-start times as each Router instance builds its query plan cache. ### References [Apollo Router 1.45.1 Release Notes](https://github.com/apollographql/router/releases/tag/v1.45.1)

### Impact Any instance of Apollo Router 1.44.0 or 1.45.0 that is using Distributed Query Plan Caching is impacted. These versions were released on 2024-04-12 and 2024-04-22 respectively. The affected versions of Apollo Router contain a bug that could lead to unexpected operations being executed, which can result in unintended data or effects. This only affects Router instances configured to use distributed query plan caching. Router versions other than the ones listed above, and all Router deployments that are not using distributed query plan caching, are unaffected by this defect. If you are using the affected versions, you can check your router’s configuration YAML to verify if you are impacted: ```yaml supergraph: query_planning: cache: # Look for this config below redis: urls: ["redis://..."] ``` A full reference on the[ Distributed Query Plan Caching feature is available here.](https://www.apollographql.com/docs/router/configuration/distributed-caching/#distributed-query-plan-caching) ### Impact detail The root cause of this defect is a bug in Apollo Router’s cache retrieval logic: When this defect is present and distributed query planning caching is enabled, asking the Router to execute an operation (whether it is a query, a mutation, or a subscription) may result in an unexpected variation of that operation being executed or the generation of unexpected errors. The issue stems from inadvertently executing a modified version of a previously executed operation, whose query plan is stored in the underlying cache (specifically, Redis). Depending on the type of the operation, the result may vary. For a query, results may be fetched that don’t match what was requested (e.g., rather than running `fetchUsers(type: ENTERPRISE)` the Router may run `fetchUsers(type: TRIAL)`. For a mutation, this may result in incorrect mutations being sent to underlying subgraph servers (e.g., rather than sending `deleteUser(id: 10)` to a subgraph, the Router may run `deleteUser(id: 12)`. ### Patches Apollo Router 1.45.1 If you are using distributed query plan caching, please either upgrade to version 1.45.1 or above or downgrade to version 1.43.2 of the Apollo Router. We do not recommend Apollo Router versions 1.44.0 or 1.45.0 for use and have withdrawn these releases. If you use impacted versions in production, we recommend that you migrate away immediately by redeploying to an unaffected Router version. For non-production use cases, we recommend you migrate at your earliest convenience. ### Workarounds If you cannot upgrade or downgrade, you can disable distributed query plan caching by removing the `supergraph.query_planning.cache.redis.urls` configuration. Please note that when distributed query plan caching is disabled, each Router instance will maintain its own in-memory query plan cache. This may increase resource utilization for each Router instance and could increase cold-start times as each Router instance builds its query plan cache. ### References [Apollo Router 1.45.1 Release Notes](https://github.com/apollographql/router/releases/tag/v1.45.1)

Apollo Router es un router de gráficos configurable escrito en Rust para ejecutar un supergrafo federado que utiliza Apollo Federation 2. Las versiones afectadas de Apollo Router contienen un error que, en circunstancias limitadas, podría provocar la ejecución de operaciones inesperadas que pueden generar datos no deseados o efectos. Esto solo afecta a las instancias del router configuradas para utilizar el almacenamiento en caché del plan de consultas distribuidas. La causa principal de este defecto es un error en la lógica de recuperación de caché del router Apollo: cuando este defecto está presente y el almacenamiento en caché de planificación de consultas distribuidas está habilitado, se solicita al router que ejecute una operación (ya sea una consulta, una mutación o una suscripción). puede resultar en una variación inesperada de esa operación que se ejecuta o en la generación de errores inesperados. El problema surge de la ejecución inadvertida de una versión modificada de una operación ejecutada previamente, cuyo plan de consulta se almacena en la memoria caché subyacente (específicamente, Redis). Dependiendo del tipo de operación, el resultado puede variar. Para una consulta, es posible que se obtengan resultados que no coincidan con lo solicitado (por ejemplo, en lugar de ejecutar `fetchUsers(type: ENTERPRISE)`, el router puede ejecutar `fetchUsers(type: TRIAL)`. Para una mutación, esto puede resultar en mutaciones incorrectas que se envían a servidores de subgrafos subyacentes (por ejemplo, en lugar de enviar `deleteUser(id: 10)` a un subgrafo, el router puede ejecutar `deleteUser(id: 12)`. Se recomienda a los usuarios que utilicen el almacenamiento en caché de planes de consulta distribuidos que actualicen a la versión 1.45.1 o superior, o bien que reduzcan a la versión 1.43.2 del router Apollo. No se recomienda el uso de las versiones 1.44.0 o 1.45.0 del router Apollo y se han retirado. Los usuarios que no puedan actualizar pueden desactivar el almacenamiento en caché del plan de consultas distribuidas para mitigar este problema.