SSH servers parsing GSSAPI authentication requests do not validate the number of mechanisms specified in the request, allowing an attacker to cause unbounded memory consumption.

Parsing a maliciously crafted DER payload could allocate large amounts of memory, causing memory exhaustion.

An attacker can cause excessive memory growth in a Go server accepting HTTP/2 requests. HTTP/2 server connections contain a cache of HTTP header keys sent by the client. While the total number of entries in this cache is capped, an attacker sending very large keys can cause the server to allocate approximately 64 MiB per open connection.

Extremely large RSA keys in certificate chains can cause a client/server to expend significant CPU time verifying signatures. With fix, the size of RSA keys transmitted during handshakes is restricted to <= 8192 bits. Based on a survey of publicly trusted RSA keys, there are currently only three certificates in circulation with keys larger than this, and all three appear to be test certificates that are not actively deployed. It is possible there are larger keys in use in private PKIs, but we target the web PKI, so causing breakage here in the interests of increasing the default safety of users of crypto/tls seems reasonable.

Despite HTTP headers having a default limit of 1MB, the number of cookies that can be parsed does not have a limit. By sending a lot of very small cookies such as "a=;", an attacker can make an HTTP server allocate a large amount of structs, causing large memory consumption.

SSH Agent servers do not validate the size of messages when processing new identity requests, which may cause the program to panic if the message is malformed due to an out of bounds read.

In archive/zip in Go before 1.16.8 and 1.17.x before 1.17.1, a crafted archive header (falsely designating that many files are present) can cause a NewReader or OpenReader panic. NOTE: this issue exists because of an incomplete fix for CVE-2021-33196.

When parsing a multipart form (either explicitly with Request.ParseMultipartForm or implicitly with Request.FormValue, Request.PostFormValue, or Request.FormFile), limits on the total size of the parsed form were not applied to the memory consumed while reading a single form line. This permits a maliciously crafted input containing very long lines to cause allocation of arbitrarily large amounts of memory, potentially leading to memory exhaustion. With fix, the ParseMultipartForm function now correctly limits the maximum size of form lines.

The processing time for parsing some invalid inputs scales non-linearly with respect to the size of the input. This affects programs which parse untrusted PEM inputs.

The net/url package does not set a limit on the number of query parameters in a query. While the maximum size of query parameters in URLs is generally limited by the maximum request header size, the net/http.Request.ParseForm method can parse large URL-encoded forms. Parsing a large form containing many unique query parameters can cause excessive memory consumption.

QUIC connections do not set an upper bound on the amount of data buffered when reading post-handshake messages, allowing a malicious QUIC connection to cause unbounded memory growth. With fix, connections now consistently reject messages larger than 65KiB in size.

A malicious HTTP/2 client which rapidly creates requests and immediately resets them can cause excessive server resource consumption. While the total number of requests is bounded by the http2.Server.MaxConcurrentStreams setting, resetting an in-progress request allows the attacker to create a new request while the existing one is still executing. With the fix applied, HTTP/2 servers now bound the number of simultaneously executing handler goroutines to the stream concurrency limit (MaxConcurrentStreams). New requests arriving when at the limit (which can only happen after the client has reset an existing, in-flight request) will be queued until a handler exits. If the request queue grows too large, the server will terminate the connection. This issue is also fixed in golang.org/x/net/http2 for users manually configuring HTTP/2. The default stream concurrency limit is 250 streams (requests) per HTTP/2 connection. This value may be adjusted using the golang.org/x/net/http2 package; see the Server.MaxConcurrentStreams setting and the ConfigureServer function.

The TIFF decoder does not place a limit on the size of compressed tile data. A maliciously-crafted image can exploit this to cause a small image (both in terms of pixel width/height, and encoded size) to make the decoder decode large amounts of compressed data, consuming excessive memory and CPU.

Multipart form parsing can consume large amounts of CPU and memory when processing form inputs containing very large numbers of parts. This stems from several causes: 1. mime/multipart.Reader.ReadForm limits the total memory a parsed multipart form can consume. ReadForm can undercount the amount of memory consumed, leading it to accept larger inputs than intended. 2. Limiting total memory does not account for increased pressure on the garbage collector from large numbers of small allocations in forms with many parts. 3. ReadForm can allocate a large number of short-lived buffers, further increasing pressure on the garbage collector. The combination of these factors can permit an attacker to cause an program that parses multipart forms to consume large amounts of CPU and memory, potentially resulting in a denial of service. This affects programs that use mime/multipart.Reader.ReadForm, as well as form parsing in the net/http package with the Request methods FormFile, FormValue, ParseMultipartForm, and PostFormValue. With fix, ReadForm now does a better job of estimating the memory consumption of parsed forms, and performs many fewer short-lived allocations. In addition, the fixed mime/multipart.Reader imposes the following limits on the size of parsed forms: 1. Forms parsed with ReadForm may contain no more than 1000 parts. This limit may be adjusted with the environment variable GODEBUG=multipartmaxparts=. 2. Form parts parsed with NextPart and NextRawPart may contain no more than 10,000 header fields. In addition, forms parsed with ReadForm may contain no more than 10,000 header fields across all parts. This limit may be adjusted with the environment variable GODEBUG=multipartmaxheaders=.

archive/zip uses a super-linear file name indexing algorithm that is invoked the first time a file in an archive is opened. This can lead to a denial of service when consuming a maliciously constructed ZIP archive.

Go before 1.10.8 and 1.11.x before 1.11.5 mishandles P-521 and P-384 elliptic curves, which allows attackers to cause a denial of service (CPU consumption) or possibly conduct ECDH private key recovery attacks.

An attacker can craft a malformed TIFF image which will consume a significant amount of memory when passed to DecodeConfig. This could lead to a denial of service.

A denial of service is possible from excessive resource consumption in net/http and mime/multipart. Multipart form parsing with mime/multipart.Reader.ReadForm can consume largely unlimited amounts of memory and disk files. This also affects form parsing in the net/http package with the Request methods FormFile, FormValue, ParseMultipartForm, and PostFormValue. ReadForm takes a maxMemory parameter, and is documented as storing "up to maxMemory bytes +10MB (reserved for non-file parts) in memory". File parts which cannot be stored in memory are stored on disk in temporary files. The unconfigurable 10MB reserved for non-file parts is excessively large and can potentially open a denial of service vector on its own. However, ReadForm did not properly account for all memory consumed by a parsed form, such as map entry overhead, part names, and MIME headers, permitting a maliciously crafted form to consume well over 10MB. In addition, ReadForm contained no limit on the number of disk files created, permitting a relatively small request body to create a large number of disk temporary files. With fix, ReadForm now properly accounts for various forms of memory overhead, and should now stay within its documented limit of 10MB + maxMemory bytes of memory consumption. Users should still be aware that this limit is high and may still be hazardous. In addition, ReadForm now creates at most one on-disk temporary file, combining multiple form parts into a single temporary file. The mime/multipart.File interface type's documentation states, "If stored on disk, the File's underlying concrete type will be an *os.File.". This is no longer the case when a form contains more than one file part, due to this coalescing of parts into a single file. The previous behavior of using distinct files for each form part may be reenabled with the environment variable GODEBUG=multipartfiles=distinct. Users should be aware that multipart.ReadForm and the http.Request methods that call it do not limit the amount of disk consumed by temporary files. Callers can limit the size of form data with http.MaxBytesReader.

Reader.Read does not set a limit on the maximum size of file headers. A maliciously crafted archive could cause Read to allocate unbounded amounts of memory, potentially causing resource exhaustion or panics. After fix, Reader.Read limits the maximum size of header blocks to 1 MiB.

Due to an allocation of resources without limits, an uncontrolled resource consumption vulnerability exists in Silicon Labs Ember ZNet SDK prior to v7.4.0.0 (delivered as part of Silicon Labs Gecko SDK v4.4.0) which may enable attackers to trigger a bus fault and crash of the device, requiring a reboot in order to rejoin the network.

zae-limiter is a rate limiting library using the token bucket algorithm. Prior to version 0.10.1, all rate limit buckets for a single entity share the same DynamoDB partition key (`namespace/ENTITY#{id}`). A high-traffic entity can exceed DynamoDB's per-partition throughput limits (~1,000 WCU/sec), causing throttling that degrades service for that entity — and potentially co-located entities in the same partition. Version 0.10.1 fixes the issue.

A lack of appropriate timeouts in GitLab Pages included in GitLab CE/EE all versions prior to 14.7.7, 14.8 prior to 14.8.5, and 14.9 prior to 14.9.2 allows an attacker to cause unlimited resource consumption.

GitLab has remediated an issue in GitLab CE/EE affecting all versions from 12.3 before 18.6.4, 18.7 before 18.7.2, and 18.8 before 18.8.2 that could have allowed an unauthenticated user to create a denial of service condition by sending repeated malformed SSH authentication requests.

GitLab has remediated an issue in GitLab CE/EE affecting versions from 18.9 before 18.9.1 that could have under certain conditions, allowed an unauthenticated user to cause denial of service by sending specially crafted requests to a CI jobs API endpoint.

Crafted zones can lead to increased resource usage and crafted CNAME chains can lead to cache poisoning in Recursor.

webtransport-go is an implementation of the WebTransport protocol. From 0.3.0 to 0.9.0, an attacker can cause excessive memory consumption in webtransport-go's session implementation by sending a WT_CLOSE_SESSION capsule containing an excessively large Application Error Message. The implementation does not enforce the draft-mandated limit of 1024 bytes on this field, allowing a peer to send an arbitrarily large message payload that is fully read and stored in memory. This allows an attacker to consume an arbitrary amount of memory. The attacker must transmit the full payload to achieve the memory consumption, but the lack of any upper bound makes large-scale attacks feasible given sufficient bandwidth. This vulnerability is fixed in 0.10.0.

CrateDB is a distributed SQL database. A high-risk vulnerability has been identified in versions prior to 5.7.2 where the TLS endpoint (port 4200) permits client-initiated renegotiation. In this scenario, an attacker can exploit this feature to repeatedly request renegotiation of security parameters during an ongoing TLS session. This flaw could lead to excessive consumption of CPU resources, resulting in potential server overload and service disruption. The vulnerability was confirmed using an openssl client where the command `R` initiates renegotiation, followed by the server confirming with `RENEGOTIATING`. This vulnerability allows an attacker to perform a denial of service attack by exhausting server CPU resources through repeated TLS renegotiations. This impacts the availability of services running on the affected server, posing a significant risk to operational stability and security. TLS 1.3 explicitly forbids renegotiation, since it closes a window of opportunity for an attack. Version 5.7.2 of CrateDB contains the fix for the issue.

Matrix Media Repo (MMR) is a highly configurable multi-homeserver media repository for Matrix. MMR before version 1.3.5 is vulnerable to unbounded disk consumption, where an unauthenticated adversary can induce it to download and cache large amounts of remote media files. MMR's typical operating environment uses S3-like storage as a backend, with file-backed store as an alternative option. Instances using a file-backed store or those which self-host an S3 storage system are therefore vulnerable to a disk fill attack. Once the disk is full, authenticated users will be unable to upload new media, resulting in denial of service. For instances configured to use a cloud-based S3 storage option, this could result in high service fees instead of a denial of service. MMR 1.3.5 introduces a new default-on "leaky bucket" rate limit to reduce the amount of data a user can request at a time. This does not fully address the issue, but does limit an unauthenticated user's ability to request large amounts of data. Operators should note that the leaky bucket implementation introduced in MMR 1.3.5 requires the IP address associated with the request to be forwarded, to avoid mistakenly applying the rate limit to the reverse proxy instead. To avoid this issue, the reverse proxy should populate the X-Forwarded-For header when sending the request to MMR. Operators who cannot update may wish to lower the maximum file size they allow and implement harsh rate limits, though this can still lead to a large amount of data to be downloaded.

Allocation of resources without limits or throttling (CWE-770) allows an unauthenticated remote attacker to cause excessive allocation (CAPEC-130) of memory and CPU via the integration of malicious IPv4 fragments, leading to a degradation in Packetbeat.

Discourse is an open source discussion platform. Versions prior to 3.5.4, 2025.11.2, 2025.12.1, and 2026.1.0 have an application level denial of service vulnerabilityin the username change functionality at try.discourse.org. The vulnerability allows attackers to cause noticeable server delays and resource exhaustion by sending large JSON payloads to the username preference endpoint PUT /u//preferences/username, resulting in degraded performance for other users and endpoints. This issue is patched in versions 3.5.4, 2025.11.2, 2025.12.1, and 2026.1.0. No known workarounds are available.

MediaWiki before 1.36.2 allows a denial of service (resource consumption because of lengthy query processing time). Visiting Special:Contributions can sometimes result in a long running SQL query because PoolCounter protection is mishandled.

Joomla! 1.03 does not restrict the number of "Search" Mambots, which allows remote attackers to cause a denial of service (resource consumption) via a large number of Search Mambots.

A potential DoS vulnerability was discovered in GitLab CE/EE starting with version 13.7. Using a malformed TIFF images was possible to trigger memory exhaustion.

REXML is an XML toolkit for Ruby. The REXML gem before 3.2.6 has a denial of service vulnerability when it parses an XML that has many `<`s in an attribute value. Those who need to parse untrusted XMLs may be impacted to this vulnerability. The REXML gem 3.2.7 or later include the patch to fix this vulnerability. As a workaround, don't parse untrusted XMLs.

A potential DOS vulnerability was discovered in GitLab CE/EE starting with version 13.7. The stripping of EXIF data from certain images resulted in high CPU usage.

quic-go is an implementation of the QUIC protocol in Go. Versions 0.56.0 and below are vulnerable to excessive memory allocation through quic-go's HTTP/3 client and server implementations by sending a QPACK-encoded HEADERS frame that decodes into a large header field section (many unique header names and/or large values). The implementation builds an http.Header (used on the http.Request and http.Response, respectively), while only enforcing limits on the size of the (QPACK-compressed) HEADERS frame, but not on the decoded header, leading to memory exhaustion. This issue is fixed in version 0.57.0.

TYPO3 is an enterprise content management system. Starting in version 9.0.0 and prior to versions 9.5.48 ELTS, 10.4.45 ELTS, 11.5.37 LTS, 12.4.15 LTS, and 13.1.1, the `ShowImageController` (`_eID tx_cms_showpic_`) lacks a cryptographic HMAC-signature on the `frame` HTTP query parameter (e.g. `/index.php?eID=tx_cms_showpic?file=3&...&frame=12345`). This allows adversaries to instruct the system to produce an arbitrary number of thumbnail images on the server side. TYPO3 versions 9.5.48 ELTS, 10.4.45 ELTS, 11.5.37 LTS, 12.4.15 LTS, 13.1.1 fix the problem described.

OpenLiteSpeed before 1.8.1 mishandles chunked encoding.

Netty is an asynchronous event-driven network application framework for rapid development of maintainable high performance protocol servers & clients. The `HttpPostRequestDecoder` can be tricked to accumulate data. While the decoder can store items on the disk if configured so, there are no limits to the number of fields the form can have, an attacher can send a chunked post consisting of many small fields that will be accumulated in the `bodyListHttpData` list. The decoder cumulates bytes in the `undecodedChunk` buffer until it can decode a field, this field can cumulate data without limits. This vulnerability is fixed in 4.1.108.Final.

AIOHTTP is an asynchronous HTTP client/server framework for asyncio and Python. In versions 3.13.2 and below, handling of chunked messages can result in excessive blocking CPU usage when receiving a large number of chunks. If an application makes use of the request.read() method in an endpoint, it may be possible for an attacker to cause the server to spend a moderate amount of blocking CPU time (e.g. 1 second) while processing the request. This could potentially lead to DoS as the server would be unable to handle other requests during that time. This issue is fixed in version 3.13.3.

An allocation of resources without limits or throttling [CWE-770] vulnerability in FortiOS versions 7.6.0, versions 7.4.4 through 7.4.0, 7.2 all versions, 7.0 all versions, 6.4 all versions may allow a remote unauthenticated attacker to prevent access to the GUI via specially crafted requests directed at specific endpoints.

An issue was discovered in Stormshield SNS before 4.2.3 (when the proxy is used). An attacker can saturate the proxy connection table. This would result in the proxy denying any new connections.

A Regular Expression Denial of Service (ReDOS) vulnerability was discovered in Color-String version 1.5.5 and below which occurs when the application is provided and checks a crafted invalid HWB string.

In some circumstances, a stale value could have been used for a global variable in WASM JIT analysis. This resulted in incorrect compilation and a potentially exploitable crash in the content process. This vulnerability affects Firefox < 116, Firefox ESR < 102.14, and Firefox ESR < 115.1.

ImageSharp is a 2D graphics API. A vulnerability discovered in the ImageSharp library, where the processing of specially crafted files can lead to excessive memory usage in the Gif decoder. The vulnerability is triggered when ImageSharp attempts to process image files that are designed to exploit this flaw. All users are advised to upgrade to v3.1.5 or v2.1.9.

Allocation of Resources Without Limits or Throttling vulnerability in Hitachi Ops Center Common Services on Linux allows DoS.This issue affects Hitachi Ops Center Common Services: before 10.9.3-00.

A regression was introduced in the Red Hat build of python-eventlet due to a change in the patch application strategy, resulting in a patch for CVE-2021-21419 not being applied for all builds of all products.

Hono is a Web application framework that provides support for any JavaScript runtime. In versions prior to 4.9.7, a flaw in the `bodyLimit` middleware could allow bypassing the configured request body size limit when conflicting HTTP headers were present. The middleware previously prioritized the `Content-Length` header even when a `Transfer-Encoding: chunked` header was also included. According to the HTTP specification, `Content-Length` must be ignored in such cases. This discrepancy could allow oversized request bodies to bypass the configured limit. Most standards-compliant runtimes and reverse proxies may reject such malformed requests with `400 Bad Request`, so the practical impact depends on the runtime and deployment environment. If body size limits are used as a safeguard against large or malicious requests, this flaw could allow attackers to send oversized request bodies. The primary risk is denial of service (DoS) due to excessive memory or CPU consumption when handling very large requests. The implementation has been updated to align with the HTTP specification, ensuring that `Transfer-Encoding` takes precedence over `Content-Length`. The issue is fixed in Hono v4.9.7, and all users should upgrade immediately.

An issue was discovered in Mattermost Server before 4.2.0, 4.1.1, and 4.0.5. It mishandles IP-based rate limiting.

node-fetch before versions 2.6.1 and 3.0.0-beta.9 did not honor the size option after following a redirect, which means that when a content size was over the limit, a FetchError would never get thrown and the process would end without failure. For most people, this fix will have a little or no impact. However, if you are relying on node-fetch to gate files above a size, the impact could be significant, for example: If you don't double-check the size of the data after fetch() has completed, your JS thread could get tied up doing work on a large file (DoS) and/or cost you money in computing.