| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| Allocation of Resources Without Limits or Throttling vulnerability in elixir-mint mint allows a malicious HTTP/2 server to exhaust memory on the client host and cause a denial of service.
Mint.HTTP2 enforces the client's max_header_list_size setting only on the compressed size of an inbound header block, while RFC 9113 section 6.5.2 defines the limit on the decoded header list. An HPACK indexed field costs one byte on the wire and decodes to a dynamic table entry of up to 4 KB, and join_cookie_headers/1 in lib/mint/http2.ex copies every cookie value of a response into one new binary. A header block under the default 256 KB wire limit therefore makes the client allocate about 1 GB for a single response, and several such responses in one delivery exhaust the memory of the process that owns the connection or of the whole VM.
This issue affects mint: from 1.1.0 before 1.10.2. |
| authentik is an open-source identity provider. Prior to 2026.2.7, 2026.5.7, and 2026.8.2, an unauthenticated attacker can submit a malformed SAML message to an authentik deployment using SAML in either the identity-provider or SAML source role. The message can stop the worker handling /application/saml/* or /source/saml/*, causing the requests assigned to that worker to fail. Worker process termination and automatic restart do not destroy database-backed sessions, but continued malicious messages can cause a sustained share of legitimate traffic to fail. Other protocol implementations are not affected. This issue is fixed in versions 2026.2.7, 2026.5.7, and 2026.8.2. |
| Klever-Go is the Go implementation of the Klever blockchain protocol. Prior to 1.7.20, the default-open GET /subscribe endpoint in network/api/websocket/routes.go accepts unauthenticated WebSocket clients with permissive origin handling, does not call SetReadLimit to bound message size, and has no live-connection cap. SocketHub.HandleClientInsertion also accepts an unbounded address list that grows addressSubscription, and client.loopIn continues reading without a size limit, allowing one client to grow subscription maps or many clients to retain goroutines, buffered channels, and descriptors. The global HTTP request throttler does not count upgraded live WebSocket connections. Because the REST and WebSocket API runs in the node process, memory or scheduler exhaustion can crash the node and interrupt P2P and consensus participation. This issue is fixed in version 1.7.20. |
| RabbitMQ is a messaging and streaming broker. Prior to versions 4.2.7 and 4.3.1, The runtime-parameters lookup path coerces the URL :component segment to an atom with rabbit_data_coercion:to_atom/1 in lookup_component/1 (deps/rabbit/src/rabbit_runtime_parameters.erl), creating a new atom for any previously unseen value. A safe helper, rabbit_registry:binary_to_type/1, which uses binary_to_existing_atom with a catch, already exists but is not used at this call site. lookup_component/1 calls rabbit_data_coercion:to_atom(Component) on the :component segment of the request URL, converting an attacker-supplied string into a new atom. Because the Erlang atom table is bounded and atoms are never garbage collected, an authorized policymaker issuing roughly one million requests with distinct component values can exhaust the atom table and crash the node, resulting in a denial of service. Preconditions include Exploitation requires policymaker privileges and roughly one million requests.. This issue is fixed in versions 4.2.7 and 4.3.1. |
| RabbitMQ is a messaging and streaming broker. Prior to versions 3.13.15, 4.0.20, 4.1.11, and 4.2.6, parse_array_primitive/2 for constructor 0x45 (list0) returns an element with byte-width B = 0. The enclosing array32 parser at line 148 reads a 4-byte Count from the wire and loops Count times consuming B bytes each , with B = 0, no input is consumed and the loop builds a list of Count empty elements bounded only by the 32-bit field. The SASL-mechanisms / SASL-init frame is parsed by amqp10_framing:decode_bin/1 from rabbit_amqp_reader.erl:412 before authentication completes. The pre-auth incoming_max_frame_size (default 8192 bytes) caps the frame, not the Count field, so a 19-byte payload with Count = 0xFFFFFFFF is accepted. No max_heap_size is set on the reader process. An unauthenticated network attacker can crash any RabbitMQ node that has the AMQP 1.0 listener enabled (default port 5672) by sending a single ~19-byte frame. The reader process attempts to build a list of ~4 billion empty elements, exhausting heap memory and terminating the Erlang VM. All tenants and protocols on the node lose service. Preconditions include Network reachability to the AMQP listener (port 5672, enabled by default) No authentication required. This issue is fixed in versions 3.13.15, 4.0.20, 4.1.11, and 4.2.6. |
| The brace-expansion library generates arbitrary strings containing a common prefix and suffix. Prior to 1.1.20, 2.1.6, 3.0.8, and 5.0.11, deeply nested brace groups cause expand_() to recurse once per nesting level at comma-member and single-set expansion sites, exhausting the native stack before output limits can apply and potentially terminating the Node.js process. expand_ performs uncontrolled recursion for nested brace alternatives and single-part sets. deeply nested brace groups supplied as an untrusted pattern. expand_ is affected. expand is affected. Comma members is affected. Single set is affected. native stack exhaustion during nested sub-expansion. process-terminating denial of service. This issue is fixed in versions 1.1.20, 2.1.6, 3.0.8, and 5.0.11. |
| This CVE ID has been rejected as a duplicate. |
| This CVE ID has been rejected as a duplicate. |
| This CVE ID has been rejected as a duplicate. |
| PyJWT is a Python implementation of JSON Web Token standards. Prior to 2.14.0, PyJWT get_signing_key_from_jwt is affected because unknown kid misses force refreshes without a negative cache or minimum refresh interval. This occurs when unauthenticated tokens repeatedly use the same unknown kid or varying kid values absent from the cached JWKS. As a result, each cache miss causes PyJWKClient to refresh the JWKS. Consequently, attacker traffic can amplify outbound requests to the configured JWKS endpoint. This issue is fixed in version 2.14.0. |
| Envoy is an open source edge and service proxy designed for cloud-native applications. Prior to 1.36.10, 1.37.6, 1.38.4, and 1.39.1, Envoy copies every decoded HTTP/2 Host header value before discarding it when :authority is already present. The discarded value bypasses saveHeader, so its bytes and count are not charged against request header limits. An unauthenticated client can use HPACK indexing to submit many references to a large Host value across a bounded number of streams, forcing extreme header-copy allocation and causing the proxy to be out-of-memory killed. The relevant scope boundary is that the demonstrated amplification uses HTTP/2 HPACK and the duplicate Host discard behavior. This issue is fixed in versions 1.36.10, 1.37.6, 1.38.4, and 1.39.1. |
| A flaw was found in the Micrometer user-event metrics listener of Keycloak, a solution for integrated identity and access management. The issue occurs when the listener is configured to include the idp tag. An unauthenticated attacker can send requests to the identity broker login endpoint using arbitrary provider aliases, causing the system to create an unlimited number of metric time series. This can lead to excessive memory consumption and degrade the performance of both the server and its monitoring tools. |
| ip-address is a library for parsing and manipulating IPv4 and IPv6 addresses in JavaScript. Prior to 10.7.1, the Address6 constructor, Address6.isValid, and parse code in src/ipv6.ts accept unbounded strings and expand invalid characters through RE_BAD_CHARACTERS into large diagnostics. Material impact occurs only when an application accepts a very large attacker-controlled field and passes it to Address6 parsing without an earlier length bound. Common URL and header limits, and common body-parser defaults, generally constrain the effect; common defaults typically exclude 32 MiB fields. Megabyte-scale fields can cause a synchronous stall and high transient memory use, approximately 16 MiB can trigger an invalid string length exception, and process termination occurs at approximately 32 MiB. The affected entry points include Address6.isValid and construction paths that reach parse. This issue is fixed in version 10.7.1. |
| Suricata is a network Intrusion Detection System, Intrusion Prevention System and Network Security Monitoring engine. Starting in version 8.0.0 and prior to version 8.0.5, when Lua rule execution is enabled, the Lua sandbox memory limit was not consistently enforced for new allocations. Certain Lua allocation patterns could exceed `security.lua.max-bytes` without triggering the intended memory limit, making the configured sandbox limit unreliable. This requires Lua rules to be enabled and an affected Lua script/rule to be loaded. Version 8.0.5 contains a fix. As a workaround, disable `security.lua.allow-rules` unless Lua rules are required. |
| vLLM is an inference and serving engine for large language models. From 0.19.0 until 0.26.0, the /v1/completions CompletionRequest.prompt field in vllm/entrypoints/openai/completion/protocol.py accepts an unbounded list[str] or list[list[int]], prompt_to_seq() in vllm/renderers/inputs/preprocess.py and OnlineRenderer.preprocess_completion() in vllm/renderers/online_renderer.py expand every element, and vllm/entrypoints/openai/completion/serving.py creates one engine generator and response slot per prompt, allowing an authenticated API client to exhaust CPU, memory, async scheduling capacity, engine request slots, and response buffering with one request. This issue is fixed in version 0.26.0. |
| Suricata is a network Intrusion Detection System, Intrusion Prevention System and Network Security Monitoring engine. From 8.0.0 until 8.0.6, the DHCP parser in rust/src/dhcp/dhcp.rs creates stateless transactions without recording their packet direction with AppLayerTxData::for_direction(), so a sensor that observes only one direction cannot mark the unseen direction inspected or free completed transactions. The RDP parser in rust/src/rdp/rdp.rs has the same direction-state defect. The per-flow transaction list can grow without bound and cleanup repeatedly scans it, causing increasing CPU and memory consumption and eventual denial of service. This issue is fixed in version 8.0.6. |
| Suricata is a network Intrusion Detection System, Intrusion Prevention System and Network Security Monitoring engine. From 8.0.5 until 8.0.6, the FTP parser in src/app-layer-ftp.c can continue allocating transactions after app-layer.protocols.ftp.max-tx is reached while processing one large chunk of FTP command data. The oversized transaction list is repeatedly processed with quadratic complexity after the too_many_transactions event, allowing crafted FTP traffic to degrade packet processing, reduce monitoring visibility, or cause denial of service. This issue is fixed in version 8.0.6. |
| Netty's HTTP/3 codec (io.netty:netty-codec-http3) from 4.2.0.Final through 4.2.17.Final retains unbounded per-stream QPACK encoder state. QpackEncoder stores a queue and a dynamic-table index tracker for every encoded field section that references the QPACK dynamic table, keyed by the peer-controlled QUIC stream ID, and these entries are released only when the remote decoder sends a Section Acknowledgment or Stream Cancellation instruction — not when the HTTP/3 stream completes. There is no limit on the number of tracked streams, field sections, or retained bytes. A remote, unauthenticated HTTP/3 client can advertise a non-zero QPACK dynamic-table capacity, acknowledge the table insertion so the server reuses a dynamically indexed response header, and then omit all mandatory Section Acknowledgments while issuing sequential requests over a single QUIC connection, bypassing concurrent-stream limits and causing unbounded heap growth until the server exhausts memory (denial of service). Fixed in 4.2.18.Final. |
| vllm before 0.29.0 fails to enforce VLLM_MAX_AUDIO_CLIP_FILESIZE_MB limit in multimodal chat audio decoding, allowing unauthenticated clients to bypass file size restrictions. Attackers can submit oversized audio files through chat endpoints to consume excessive memory and CPU resources during decoding. |
| GitLab has remediated an issue in GitLab CE/EE affecting all versions from 18.4.6 before 19.1.8, 19.2 before 19.2.6, and 19.3 before 19.3.2 that under certain conditions could have allowed an unauthenticated user to cause denial of service due to improper resource allocation limits in the GraphQL complexity calculation logic. |
