# US 2021/0240640 A1 · Using unikernels to design high performance integration workflows > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2021/0240640 A1 - Document ID: US20210240640A1 - Status: Published application - Kind: A1 - Application number: 16/777,062 - Filed: 2020-01-30 - Published: 2021-08-05 - Earliest claimed: 2020-01-30 - Assignees: Sap SE - Applicants: Sap SE - Inventors: Shashank Mohan Jain - Web page: https://patentgrep.com/patent/US20210240640A1 ## CPC - G06F 13/122: {where hardware performs an I/O function other than control of data transfer} - G02B 6/29382: {including at least adding or dropping a signal, i.e. passing the majority of signals} - G06F 9/505: {considering the load} - G06F 9/545: {where tasks reside in different layers, e.g. user- and kernel-space} - G06N 20/00: Machine learning - H04L 67/02: based on web technology, e.g. hypertext transfer protocol [HTTP] - H04L 67/568: Storing data temporarily at an intermediate stage, e.g. caching - H04L 69/16: Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP] - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G02B: OPTICAL ELEMENTS, SYSTEMS OR APPARATUS - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) ## Abstract According to some embodiments, an integration system may include a source data store that contains a plurality of workloads (e.g., data packets). A processing platform may retrieve a workload from the source data store via first unikernel-based workload processing. The processing platform may then process the workload (e.g., filtering or dropping) to generate an output result and arrange for the output result to be provided to a sink destination via second unikernel-based workload processing (e.g., associated with a Linux event loop model of input output multiplexing). In some embodiments, the processing platform initially evaluates the workload in the source data store to determine if the workload meets a predetermined condition. If the workload does not meet the predetermined condition, the retrieving and arranging may be performed via container-based workload processing instead of unikernel-based workload processing. ## Claims ### Claim 1 An integration system, comprising: a source data store containing a plurality of workloads; a processing platform, coupled to the source data store, including: a computer processor, and a memory storage device including instructions that when executed by the computer processor enable the integration system to: retrieve a workload from the source data store via first unikernel-based workload processing, process the workload to generate an output result, and arrange for the output result to be provided to a sink destination via second unikernel-based workload processing. ### Claim 2 (depends on 1) The integration system of claim 1, wherein the workload is associated with an integration workflow. ### Claim 3 (depends on 2) The integration system of claim 2, wherein the workload is associated with an input output packet accepted at a network stack. ### Claim 4 (depends on 2) The integration system of claim 2, wherein the source data store is associated with a first Hypertext Transfer Protocol (“HTTP”) endpoint and the sink destination is associated with a second HTTP endpoint. ### Claim 5 (depends on 1) The integration system of claim 1, wherein the processing performed by the processing platform is associated with at least one of: (i) filtering, and (ii) dropping. ### Claim 6 (depends on 1) The integration system of claim 1, wherein the unikernel-based workload processing is associated with a Linux event loop model of input output multiplexing. ### Claim 7 (depends on 6) The integration system of claim 6, wherein the processing platform uses a per tenant approach to build appropriate security primitives associated with the workload. ### Claim 8 (depends on 1) The integration system of claim 1, wherein the computer processor of the processing platform initially evaluates a set of workloads in the source data store to determine if the set of workloads meets a predetermined condition and, if the workload does not meet the predetermined condition, said retrieving and arranging are performed via container-based workload processing instead of unikernel-based workload processing. ### Claim 9 (depends on 8) The integration system of claim 8, wherein the evaluation is based on a machine learning algorithm. ### Claim 10 A computer-implemented method, comprising: evaluating, by a computer processor of a processing platform, a workload in a source data store to determine if the workload meets a predetermined condition; if the computer processor of the processing platform determines that the workload does not meet the predetermined condition: retrieving, by the computer processor of the processing platform, the workload from the source data store via first container-based workload processing, processing, by the computer processor of the processing platform, the workload to generate an output result, and arranging, by the computer processor of the processing platform, for the output result to be provided to a sink destination via second container-based workload processing; and if the computer processor of the processing platform determines that the workload meets the predetermined condition: retrieving, by the computer processor of the processing platform, the workload from the source data store via first unikernel-based workload processing, processing, by the computer processor of the processing platform, the workload to generate an output result, and arranging, by the computer processor of the processing platform, for the output result to be provided to the sink destination via second unikernel-based workload processing. ### Claim 11 (depends on 10) The method of claim 10, wherein the workload is associated with an integration workflow. ### Claim 12 (depends on 11) The method of claim 11, wherein the workload is associated with an input output packet accepted at a network stack. ### Claim 13 (depends on 11) The method of claim 11, wherein the source data store is associated with a first Hypertext Transfer Protocol (“HTTP”) endpoint and the sink destination is associated with a second HTTP endpoint. ### Claim 14 (depends on 10) The method of claim 10, wherein the processing performed by the processing platform is associated with at least one of: (i) filtering, and (ii) dropping. ### Claim 15 (depends on 10) The method of claim 10, wherein the unikernel-based workload processing is associated with a Linux event loop model of input output multiplexing. ### Claim 16 (depends on 15) The method of claim 15, wherein the processing platform uses a per tenant approach to build appropriate security primitives associated with the workload. ### Claim 17 A non-transitory, computer-readable medium storing instructions that, when executed by a computer processor, cause the computer processor to perform a method, the method comprising: evaluating, by a machine learning algorithm a computer processor of a processing platform, a set of workloads in a source data store to determine if the set of workloads meets a predetermined condition; if the computer processor of the processing platform determines that the set of workloads does not meet the predetermined condition: retrieving, by the computer processor of the processing platform, a workload from the source data store via first container-based workload processing, processing, by the computer processor of the processing platform, the workload to generate an output result, and arranging, by the computer processor of the processing platform, for the output result to be provided to a sink destination via second container-based workload processing; and if the computer processor of the processing platform determines that the set of workloads meets the predetermined condition: retrieving, by the computer processor of the processing platform, the workload from the source data store via first unikernel-based workload processing, processing, by the computer processor of the processing platform, the workload to generate an output result, and arranging, by the computer processor of the processing platform, for the output result to be provided to the sink destination via second unikernel-based workload processing. ### Claim 18 (depends on 17) The medium of claim 17, wherein the workload is associated with an integration workflow. ### Claim 19 (depends on 18) The medium of claim 18, wherein the workload is associated with an input output packet accepted at a network stack. ### Claim 20 (depends on 18) The medium of claim 18, wherein the source data store is associated with a first Hypertext Transfer Protocol (“HTTP”) endpoint and the sink destination is associated with a second HTTP endpoint.