@elastic/elasticsearch

Uncontrolled Resource Consumption (CWE-400) in Elasticsearch can lead to a denial of service via Excessive Allocation (CAPEC-130). An authenticated user can submit a specially crafted bulk request that causes sustained high CPU consumption, which can render the affected node unable to process requests.

Allocation of Resources Without Limits or Throttling (CWE-770) in Elasticsearch can lead to a denial of service via Excessive Allocation (CAPEC-130). A user with elevated privileges can submit a specially crafted machine learning request that causes excessive memory consumption, which may render the affected node unavailable.

Uncontrolled Recursion (CWE-674) in Elasticsearch can lead to a denial of service via Excessive Allocation (CAPEC-130). An authenticated user can submit a specially crafted query that causes excessive resource consumption while the request is processed, which may render the affected node unavailable.

Allocation of Resources Without Limits or Throttling (CWE-770) in Elasticsearch can allow an authenticated user with snapshot restore privileges to cause Excessive Allocation (CAPEC-130) of memory and a denial of service (DoS) via crafted HTTP request.

Allocation of Resources Without Limits or Throttling (CWE-770) in Elasticsearch can allow a low-privileged authenticated user to cause Excessive Allocation (CAPEC-130) causing a persistent denial of service (OOM crash) via submission of oversized user settings data.

Improper Authentication in Elasticsearch PKI realm can lead to user impersonation via specially crafted client certificates. A malicious actor would need to have such a crafted client certificate signed by a legitimate, trusted Certificate Authority.

Insertion of sensitive information in log file in Elasticsearch can lead to loss of confidentiality under specific preconditions when auditing requests to the reindex API https://www.elastic.co/docs/api/doc/elasticsearch/operation/operation-reindex

Uncontrolled Resource Consumption in Elasticsearch while evaluating specifically crafted search templates with Mustache functions can lead to Denial of Service by causing the Elasticsearch node to crash.

An issue was discovered in Elasticsearch, where a large recursion using the Well-KnownText formatted string with nested GeometryCollection objects could cause a stackoverflow.

A flaw was discovered in Elasticsearch, where a large recursion using the innerForbidCircularReferences function of the PatternBank class could cause the Elasticsearch node to crash. A successful attack requires a malicious user to have read_pipeline Elasticsearch cluster privilege assigned to them.

An allocation of resources without limits or throttling in Elasticsearch can lead to an OutOfMemoryError exception resulting in a crash via a specially crafted query using an SQL function.

An issue was discovered where improper authorization controls affected certain queries that could allow a malicious actor to circumvent Document Level Security in Elasticsearch and get access to documents that their roles would normally not allow.

It was discovered by Elastic engineering that when elasticsearch-certutil CLI tool is used with the csr option in order to create a new Certificate Signing Requests, the associated private key that is generated is stored on disk unencrypted even if the --pass parameter is passed in the command invocation.

An issue was discovered by Elastic whereby Watcher search input logged the search query results on DEBUG log level. This could lead to raw contents of documents stored in Elasticsearch to be printed in logs. Elastic has released 8.11.2 and 7.17.16 that resolves this issue by removing this excessive logging. This issue only affects users that use Watcher and have a Watch defined that uses the search input and additionally have set the search input’s logger to DEBUG or finer, for example using: org.elasticsearch.xpack.watcher.input.search, org.elasticsearch.xpack.watcher.input, org.elasticsearch.xpack.watcher, or wider, since the loggers are hierarchical.

A flaw was discovered in Elasticsearch, affecting document ingestion when an index template contains a dynamic field mapping of “passthrough” type. Under certain circumstances, ingesting documents in this index would cause a StackOverflow exception to be thrown and ultimately lead to a Denial of Service. Note that passthrough fields is an experimental feature.

It was identified that if a cross-cluster API key https://www.elastic.co/guide/en/elasticsearch/reference/8.14/security-api-create-cross-cluster-api-key.html#security-api-create-cross-cluster-api-key-request-body  restricts search for a given index using the query or the field_security parameter, and the same cross-cluster API key also grants replication for the same index, the search restrictions are not enforced during cross cluster search operations and search results may include documents and terms that should not be returned. This issue only affects the API key based security model for remote clusters https://www.elastic.co/guide/en/elasticsearch/reference/8.14/remote-clusters.html#remote-clusters-security-models  that was previously a beta feature and is released as GA with 8.14.0

An uncaught exception in Elasticsearch >= 8.4.0 and < 8.11.1 occurs when an encrypted PDF is passed to an attachment processor through the REST API. The Elasticsearch ingest node that attempts to parse the PDF file will crash. This does not happen with password-protected PDF files or with unencrypted PDF files.

Incorrect Authorization issue exists in the API key based security model for Remote Cluster Security, which is currently in Beta, in Elasticsearch 8.10.0 and before 8.13.0. This allows a malicious user with a valid API key for a remote cluster configured to use the new Remote Cluster Security to read arbitrary documents from any index on the remote cluster, and only if they use the Elasticsearch custom transport protocol to issue requests with the target index ID, the shard ID and the document ID. None of Elasticsearch REST API endpoints are affected by this issue.

A flaw was discovered in Elasticsearch, where processing a document in a deeply nested pipeline on an ingest node could cause the Elasticsearch node to crash.

An issue was identified that allowed the unsafe deserialization of java objects from hadoop or spark configuration properties that could have been modified by authenticated users. Elastic would like to thank Yakov Shafranovich, with Amazon Web Services for reporting this issue.

It was identified that malformed scripts used in the script processor of an Ingest Pipeline could cause an Elasticsearch node to crash when calling the Simulate Pipeline API.

An issue was found with how API keys are created with the Fleet-Server service account. When an API key is created with a service account, it is possible that the API key could be created with higher privileges than intended. Using this vulnerability, a compromised Fleet-Server service account could escalate themselves to a super-user.

An issue has been identified with how Elasticsearch handled incoming requests on the HTTP layer. An unauthenticated user could force an Elasticsearch node to exit with an OutOfMemory error by sending a moderate number of malformed HTTP requests. The issue was identified by Elastic Engineering and we have no indication that the issue is known or that it is being exploited in the wild.

A flaw was discovered in Elasticsearch, affecting the _search API that allowed a specially crafted query string to cause a Stack Overflow and ultimately a Denial of Service.

Elasticsearch generally filters out sensitive information and credentials before logging to the audit log. It was found that this filtering was not applied when requests to Elasticsearch use certain deprecated URIs for APIs. The impact of this flaw is that sensitive information such as passwords and tokens might be printed in cleartext in Elasticsearch audit logs. Note that audit logging is disabled by default and needs to be explicitly enabled and even when audit logging is enabled, request bodies that could contain sensitive information are not printed to the audit log unless explicitly configured.

A Denial of Service flaw was discovered in Elasticsearch. Using this vulnerability, an unauthenticated attacker could forcibly shut down an Elasticsearch node with a specifically formatted network request.

A flaw was discovered in Elasticsearch 7.17.0’s upgrade assistant, in which upgrading from version 6.x to 7.x would disable the in-built protections on the security index, allowing authenticated users with “*” index permissions access to this index.

Elasticsearch before 7.14.0 did not apply document and field level security to searchable snapshots. This could lead to an authenticated user gaining access to information that they are unauthorized to view.

In Elasticsearch versions before 7.13.3 and 6.8.17 an uncontrolled recursion vulnerability that could lead to a denial of service attack was identified in the Elasticsearch Grok parser. A user with the ability to submit arbitrary queries to Elasticsearch could create a malicious Grok query that will crash the Elasticsearch node.

All versions of Elastic Cloud Enterprise has the Elasticsearch “anonymous” user enabled by default in deployed clusters. While in the default setting the anonymous user has no permissions and is unable to successfully query any Elasticsearch APIs, an attacker could leverage the anonymous user to gain insight into certain details of a deployed cluster.

A memory disclosure vulnerability was identified in Elasticsearch 7.10.0 to 7.13.3 error reporting. A user with the ability to submit arbitrary queries to Elasticsearch could submit a malformed query that would result in an error message returned containing previously used portions of a data buffer. This buffer could contain sensitive information such as Elasticsearch documents or authentication details.

In Elasticsearch versions before 7.11.2 and 6.8.15 a document disclosure flaw was found when Document or Field Level Security is used. Search queries do not properly preserve security permissions when executing certain cross-cluster search queries. This could result in the search disclosing the existence of documents the attacker should not be able to view. This could result in an attacker gaining additional insight into potentially sensitive indices.

Elasticsearch versions before 7.11.2 and 6.8.15 contain a document disclosure flaw was found in the Elasticsearch suggester and profile API when Document and Field Level Security are enabled. The suggester and profile API are normally disabled for an index when document level security is enabled on the index. Certain queries are able to enable the profiler and suggester which could lead to disclosing the existence of documents and fields the attacker should not be able to view.

A document disclosure flaw was found in Elasticsearch versions after 7.6.0 and before 7.11.0 when Document or Field Level Security is used. Get requests do not properly apply security permissions when executing a query against a recently updated document. This affects documents that have been updated and not yet refreshed in the index. This could result in the search disclosing the existence of documents and fields the attacker should not be able to view.

Elasticsearch versions before 7.10.0 and 6.8.14 have an information disclosure issue when audit logging and the emit_request_body option is enabled. The Elasticsearch audit log could contain sensitive information such as password hashes or authentication tokens. This could allow an Elasticsearch administrator to view these details.

Elasticsearch versions 7.7.0 to 7.10.1 contain an information disclosure flaw in the async search API. Users who execute an async search will improperly store the HTTP headers. An Elasticsearch user with the ability to read the .tasks index could obtain sensitive request headers of other users in the cluster. This issue is fixed in Elasticsearch 7.10.2

Elasticsearch versions before 6.8.13 and 7.9.2 contain a document disclosure flaw when Document or Field Level Security is used. Search queries do not properly preserve security permissions when executing certain complex queries. This could result in the search disclosing the existence of documents the attacker should not be able to view. This could result in an attacker gaining additional insight into potentially sensitive indices.

In Elasticsearch before 7.9.0 and 6.8.12 a field disclosure flaw was found when running a scrolling search with Field Level Security. If a user runs the same query another more privileged user recently ran, the scrolling search can leak fields that should be hidden. This could result in an attacker gaining additional permissions against a restricted index.

The fix for CVE-2020-7009 was found to be incomplete. Elasticsearch versions from 6.7.0 to 6.8.7 and 7.0.0 to 7.6.1 contain a privilege escalation flaw if an attacker is able to create API keys and also authentication tokens. An attacker who is able to generate an API key and an authentication token can perform a series of steps that result in an authentication token being generated with elevated privileges.

Elasticsearch versions from 6.7.0 before 6.8.8 and 7.0.0 before 7.6.2 contain a privilege escalation flaw if an attacker is able to create API keys. An attacker who is able to generate an API key can perform a series of steps that result in an API key being generated with elevated privileges.

Elasticsearch versions 7.0.0-7.3.2 and 6.7.0-6.8.3 contain a username disclosure flaw was found in the API Key service. An unauthenticated attacker could send a specially crafted request and determine if a username exists in the Elasticsearch native realm.

A race condition flaw was found in the response headers Elasticsearch versions before 7.2.1 and 6.8.2 returns to a request. On a system with multiple users submitting requests, it could be possible for an attacker to gain access to response header containing sensitive data from another user.

A permission issue was found in Elasticsearch versions before 5.6.15 and 6.6.1 when Field Level Security and Document Level Security are disabled and the _aliases, _shrink, or _split endpoints are used . If the elasticsearch.yml file has xpack.security.dls_fls.enabled set to false, certain permission checks are skipped when users perform one of the actions mentioned above, to make existing data available under a new index/alias name. This could result in an attacker gaining additional permissions against a restricted index.

Elasticsearch Security versions 6.4.0 to 6.4.2 contain an error in the way request headers are applied to requests when using the Active Directory, LDAP, Native, or File realms. A request may receive headers intended for another request if the same username is being authenticated concurrently; when used with run as, this can result in the request running as the incorrect user. This could allow a user to access information that they should not have access to.

Elasticsearch Security versions 6.5.0 and 6.5.1 contain an XXE flaw in Machine Learning's find_file_structure API. If a policy allowing external network access has been added to Elasticsearch's Java Security Manager then an attacker could send a specially crafted request capable of leaking content of local files on the Elasticsearch node. This could allow a user to access information that they should not have access to.