Abstract: Streaming improvements are provided by using computing, caching, and/or other resources that are provided at different network edges in a Multi-access Edge Computing (“MEC”) enhanced network. A MEC device, at a network edge, may customize a stream passing to a client device via the network edge by identifying a break in an original stream manifest passing to the client device, and by inserting a link for the client device to request and play content that is selected by the MEC device during the break. The MEC device then passes the modified stream manifest, instead of the original stream manifest, to the client device. The MEC device may also be triggered, via a prioritization message from a host in the network, to prefetch, cache, and serve a stream from the network edge to a client device accessing the network via that network edge.

Abstract: Provided is an intelligent prefetcher for prefetching media content segments following a non-sequential naming convention. The prefetching is initiated in response to a server receiving a user request for a particular segment that is not cached. The server passes the request to an origin. The origin scans metadata of a file from which the particular segment is produced. The scanning identifies time or frame offsets for subsequent segments to be prefetched. The origin constructs the requested segment and includes in the segment header, identifiers corresponding to the offset of each subsequent segment selected for prefetching. The segment passes to the server. The server detects the modified header and generates prefetch requests based on each identifier in the modified header. The server issues the prefetch request to the origin and caches the prefetched segments returned by the origin for distribution in response to subsequent user requests for the prefetched segments.

Abstract: An Edge Compute Network (“ECN”) may provide compute devices with configurable compute resources at different network edges for selective localized and low latency execution of user-defined operations. A user may provide a set of operations in any of several supported programming languages, and the ECN may create an executable environment that can execute the set of operations using the compute resources of any ECN compute device. A function manager for a compute device may selectively initialize, based on requests to execute the set of operations and an initialization policy, the executable environment so that the set of operations can be directly executed using the compute resources of that compute device upon request. The function manager may also selectively remove initialized executable environments based on compute device resource utilization, executable environment usage, and a removal policy.

Abstract: The embodiments provide peer cache filling. The peer cache filling allocates a set of caching servers to distribute content in response to user requests with a limited first subset of the set of servers having access to retrieve the content from an origin and with a larger second subset of the set of servers retrieving the content from the first subset of servers without accessing the origin. The peer cache filling dynamically escalates and deescalates the allocation of the caching servers to the first and second subsets as demand for the content rises and falls. Peer cache filling is implemented by modifying request headers to identify designated hot content, provide a request identifier hash result for identifying the ordering of servers, and provide a value for designating which servers in the ordering as primary server with access to the origin.

Abstract: Disclosed are different implementations for rapid configuration propagation across multiple servers of a distributed platform. One implementation is a push based distribution of update segments that are generated from a onetime differential analysis of an updated particular configuration relative to a previous instance of the particular configuration. Sequence numbers attached to the updated segments identify is a server's copy of a configuration is up-to-date and can receive a new updated segment or if missing intervening segments are to be retrieved from peers and applied prior to applying the new updated segment. Another implementation is a pull based distribution of compressed images of the configurations. A complete set of configurations are distributed as a compressed file system that is loaded into server memory. Individual configurations are read out of the file system and loaded into memory when implicated by client requests.

Abstract: Some embodiments provide a multi-tenant over-the-top multicast solution that integrates the per user stream customizability of unicast with the large scale streaming efficiencies of multicast. The solution involves an application, different multicast groups streaming an event with different customizations, and a manifest file or metadata identifying the different groups and customizations. The solution leverages the different multicast groups in order to provide different time shifts in the event stream, different quality level encodings of the event stream, and different secondary content to be included with a primary content stream. The application configured with the manifest file or metadata dynamically switches between the groups in order to customize the experience for a user or user device on which the application executes. Switching from multicast to unicast is also supported to supplement available customizations and for failover.

Abstract: The progressive delivery of targeted secondary content may include determining that primary content requested from a particular distribution point of a distributed platform includes a first tag to integrate targeted secondary content as part of an initial presentation of the primary content, and a second tag to integrate targeted secondary content outside the initial presentation. The progressive delivery may further include modifying the primary content by replacing the first tag so that the primary content and targeted secondary content for the initial presentation (e.g., third-party content associated with the first tag) are provided from the particular distribution point without accessing a third-party content provider, while targeted secondary content outside the initial presentation (e.g., third-party content associated with the second tag) may be selected and provided by one or more third-party content providers as originally specified in the primary content.

Abstract: Localized and global detection and mitigation of network attacks in a distributed platform are provided. The localized detection identifies attacks occurring at individual nodes of the distributed platform based on packet analysis conducted by each individual node. The global detection identifies attacks occurring across the distributed platform based on packet analysis conducted on traffic aggregated from across the distributed platform. Either detection involves inspecting headers of the sampled packets. Each header property is scored based on an amount of deviation from threshold values. The sum of scores identifies the header properties that form an attack signature. Attack protections are implemented against subsequently arriving packets with header properties matching the attack signature.

Abstract: Disclosed is a device for configuring and implementing network security for a connected network node, and for shifting the network security closer to the attack point of origin. In particular, the device may activate attack protections on different Multi-Access Edge Computing (“MEC”) devices that are physically located near or at the attack point of origin. The device may detect an attack signature based on one or more received data packets, and may provide a response with an extended header field, the attack signature, and/or other attack protection instructions. The responses may be passed to an address of a suspected attacker. MEC devices along the network path may detect and receive the responses, and implement attack protections in response. The responses may also be passed to a multicast or broadcast address that the MEC device may use to receive responses.

Abstract: An adaptive stream segment prefetcher changes the number of segments it prefetches following a client requested segment of the same stream based on conditions associated with that stream at prefetch time. The adaptive prefetcher increases or decreases the number of segments to prefetch for a particular stream based on the number of active or concurrent clients requesting that particular stream, based on the playback duration of the particular stream by one or more clients, or some combination of both. The adaptive prefetcher continuously monitors the conditions associated with the stream such that number of segments prefetched at a first time are greater or less than the number of segments prefetched at a later second time.

Abstract: Gateway devices at different sites of a primary Anycast network provide access to the sites by advertising a first set of Anycast addresses. A secondary shadow Anycast network advertises different second sets of Anycast addresses from the different sites in order to predetermine traffic shifts that occur as a result of changing one or more of the second set of Anycast addresses that are advertised from one or more of the sites. A traffic shifting device may implement a predetermined traffic shift in the primary Anycast network by selecting a particular second set of network addresses that produces a traffic shift at least equal to the predetermined traffic shift, mapping the particular second set of network addresses to a modified first set of addresses, and modifying routing in the primary Anycast network by advertising the modified first set of addresses instead of the first set of addresses.

Abstract: Dynamic and selective compression for content caching is provided for improving content delivery over a digital network. The dynamic and selective compression increased server cache size for higher cache-hit ratios that offset delays associated with compressing and decompressing content. The dynamic compression involves switching between an available set of compression tools in order to compress different files with the compression tool that is optimal for that file. The selective compression involves selectively compressing the content or files with the optimal compression tool when at least a threshold amount of space savings is obtained in an acceptable amount of time. Thus, the caching server caches compressed copies of a first set of files compressed with a first compression tool, compressed copies of a second set of files compressed with a different second compression tool, and an uncompressed third set of files.

Abstract: Some embodiments perform probabilistic request routing in addition to or instead of deterministic request routing. The probabilistic request routing is based on probabilistic models that predict the type of content being requested based on commonality in elements between different requests directed to the same type. The probabilistic models accurately route requests that have not been previously encountered and accurately route requests for content whose type is not previously known. The requests are routed across different subsets of servers that are optimized or configured for the predicted type. The probabilistic models can be defined using a decision tree. Machine learning generates and maintains the decision tree. Accuracy predicted by the different branches of the tree is updated through tracking the type of content passed in response to different routed requests. The tree structure is modified based on timestamps associated with the tree elements and based on newly encountered request elements.

Abstract: Provided is an intelligent prefetcher for prefetching media content segments following a non-sequential naming convention. The prefetching is initiated in response to a server receiving a user request for a particular segment that is not cached. The server passes the request to an origin. The origin scans metadata of a file from which the particular segment is produced. The scanning identifies time or frame offsets for subsequent segments to be prefetched. The origin constructs the requested segment and includes in the segment header, identifiers corresponding to the offset of each subsequent segment selected for prefetching. The segment passes to the server. The server detects the modified header and generates prefetch requests based on each identifier in the modified header. The server issues the prefetch request to the origin and caches the prefetched segments returned by the origin for distribution in response to subsequent user requests for the prefetched segments.

Abstract: Multiple hit load balancing provides a quasi-persistent request distribution for encrypted requests passing over secure connections as well as for multiple requests passing over the same connection. The multiple hit load balancing involves tracking object demand at each server of a set of servers. The multiple hit load balancing further involves dynamically scaling the servers that cache and directly serve frequently requested objects based on the demand that is tracked by each of the servers. For infrequently requested objects, the servers perform a peer retrieval of the objects so to limit the number of the same object being redundantly cached by multiple servers of the set of servers.

Abstract: A two-tier Anycast addressing hierarchy enables deterministic traffic management in an Anycast system. Different locations of the Anycast system advertise a common Anycast cover route comprising a first-tier Anycast address with a first address prefix, and different specific routes comprising a set of second-tier Anycast addresses with a second address prefix, wherein each address of the set of second-tier Anycast addresses falls within the already advertised Anycast cover route, and wherein the second prefix is larger than the first prefix. Domain Name System (DNS) operation can resolve queries to the cover route when normal Anycast operation is desired or can be modified to resolve a subset of queries to a particular second-tier address in order to shift a deterministic amount of traffic based on the resolved subset of queries to a deterministically selected Anycast system location based on the location from which the particular second-tier address is advertised.

Abstract: The embodiments provide systems and methods for efficiently and accurately differentiating requests directed to uncacheable content from requests directed to cacheable content based on identifiers from the requests. The differentiation occurs without analysis or retrieval of the content being requested. Some embodiments hash identifiers of prior requests that resulted in uncacheable content being served in order to set indices within a bloom filter. The bloom filter then tracks prior uncacheable requests without storing each of the identifiers so that subsequent requests for uncacheable requests can be easily identified based on a hash of the request identifier and set indices of the bloom filter. Some embodiments produce a predictive model identifying uncacheable content requests by tracking various characteristics found in identifiers of prior requests that resulted in uncacheable content being served.

Abstract: Asynchronous file tracking may include a first process that adds files to a cache and that generates different instances of a tracking file to track the files as they are entered into the cache. A second process, executing on the device, asynchronously accesses one or more instances of the tracking file at a different rate than the first process generates the tracking file instances. The second process may update a record of cached files based on a set of entries from each of the different instances of the tracking file accessed by the second process. Each set of entries may identify a different set of files that are cached by the device. The second process may then purge one or more cached files that satisfy eviction criteria while the first process continues to asynchronously add files to the cache and create new instances to track the newly cached files.

Abstract: Streaming improvements are provided by using computing, caching, and/or other resources that are provided at different network edges in a Multi-access Edge Computing (“MEC”) enhanced network. A MEC device, at a network edge, may customize a stream passing to a client device via the network edge by identifying a break in an original stream manifest passing to the client device, and by inserting a link for the client device to request and play content that is selected by the MEC device during the break. The MEC device then passes the modified stream manifest, instead of the original stream manifest, to the client device. The MEC device may also be triggered, via a prioritization message from a host in the network, to prefetch, cache, and serve a stream from the network edge to a client device accessing the network via that network edge.

Abstract: Optimized message flooding is provided for time sensitive or real-time contacting of nodes across a distributed platform. The propagation of messaging across the distributed platform is accelerated based on a parallel fan-out distribution of the flood messaging from increasing numbers of nodes with the distribution occurring over existing connections maintained between different sets of nodes. Responses are propagated via a reverse traversal of the nodes. The reverse traversal is optimized by merging responses at different tiers. A set of triggers set at the different nodes specify conditions with which to identify and preempt the slowest nodes of a lower tier from withholding message flooding completion. As such, desired messaging and responses can be collected from across the distributed platform in a determinate amount of time with minimal overhead.