Provisioning using pre-fetched data in serverless computing environments

Patent metadata

Field	Value
Publication number	US-2020396311-A1
Application number	US-202017008520-A
Country	US
Kind code	A1
Filing date	Aug 31, 2020
Priority date	Nov 30, 2017
Publication date	Dec 17, 2020
Grant date	—

Abstract

A method for data provisioning a serverless computing cluster. A plurality of user defined functions (UDFs) are received for execution on worker nodes of the serverless computing cluster. For a first UDF, one or more data locations of UDF data needed to execute the first UDF are determined. At a master node of the serverless computing cluster, a plurality of worker node tickets are received, each ticket indicating a resource availability of a corresponding worker node. The one or more data locations and the plurality of worker node tickets are analyzed to determine eligible worker nodes capable of executing the first UDF. The master node transmits a pre-fetch command to one or more of the eligible worker nodes, causing the eligible worker nodes to become a provisioned worker node for the first UDF by storing a pre-fetched first UDF data before the first UDF is assigned for execution.

First claim

We claim:

1 . A method comprising:
determining a subset of worker nodes of a plurality of work nodes eligible to execute a user defined function (UDF); generating and transmitting one or more instructions to the subset of worker nodes to pre-fetch data associated with the UDF, wherein the one or more instructions cause the subset of worker nodes to retrieve data associated with the UDF; and assigning the UDF to the subset of worker nodes.

2 . The method of claim 1 ,
wherein,
the subset worker nodes are in an idle state,
the subset worker nodes each store the data; and
the assigning of the UDF includes selecting at least one of the subset worker nodes to execute the UDF.

3 . The method of claim 1 , wherein the subset of worker nodes are selected based on at least one of a distance to a location of the data, an available capacity, or an estimated time to complete current tasks.

4 . The method of claim 1 , wherein the assigning of the UDF includes additional worker nodes of the plurality of worker nodes that have not stored the data.

5 . The method of claim 1 , further comprising:
determining whether the UDF is eligible for parallel execution; in response to determining the USF is eligible for parallel execution, generating a plurality of instructions to the subset of worker nodes, each instruction of the plurality of instructions to pre-fetch a portion of the data associated with the UDF; transmitting the plurality of instructions to the subset of work nodes, wherein each instruction of the plurality of instructions is not transmitted to more than one node of the subset of worker nodes; and assigning the UDF to the subset of worker nodes.

6 . The method of claim 1 , wherein the assigning the UDF to the subset of worker nodes is based on one or more parameters including completion time, degree of parallel computation desired, or power consumption.

7 . The method of claim 1 , wherein the one or more instructions includes at least a location of the data.

8 . A system comprising:
at least one processor; and at least one memory storing instructions, which when executed by the at least one processor, causes the at least one processor to:
determine a subset of worker nodes of a plurality of work nodes eligible to execute a user defined function (UDF);
generate and transmit one or more instructions to the subset of worker nodes to pre-fetch data associated with the UDF, wherein the one or more instructions cause the subset of worker nodes to retrieve data associated with the UDF; and
assign the UDF to the subset of worker nodes.

9 . The system of claim 8 ,
wherein,
the subset worker nodes are in an idle state,
the subset worker nodes each store the data; and
the assignment the UDF includes a selection of at least one of the subset worker nodes to execute the UDF.

10 . The system of claim 8 , wherein the subset of worker nodes are selected based on at least one of a distance to a location of the data, an available capacity, or an estimated time to complete current tasks.

11 . The system of claim 8 , wherein the assignment of the UDF includes additional worker nodes of the plurality of worker nodes that have not stored the data.

12 . The system of claim 8 , further comprising instructions, which when executed by the at least one processor causes the at least one processor to:
determine whether the UDF is eligible for parallel execution; in response to determination that the USF is eligible for parallel execution, generate a plurality of instructions to the subset of worker nodes, each instruction of the plurality of instructions to pre-fetch a portion of the data associated with the UDF; transmit the plurality of instructions to the subset of work nodes, wherein each instruction of the plurality of instructions is not transmitted to more than one node of the subset of worker nodes; and assign the UDF the subset of worker nodes.

13 . The system of claim 8 , wherein the assignment of the UDF to the subset of worker nodes is based on one or more parameters including completion time, degree of parallel computation desired, or power consumption.

14 . The system of claim 8 , wherein the one or more instructions includes at least a location of the data.

15 . At least one non-transitory computer readable medium storing instructions, which when executed by the at least one processor, causes the at least one processor to:
determine a subset of worker nodes of a plurality of work nodes eligible to execute a user defined function (UDF); generate and transmit one or more instructions to the subset of worker nodes to pre-fetch data associated with the UDF, wherein the one or more instructions cause the subset of worker nodes to retrieve data associated with the UDF; and assign the UDF to the subset of worker nodes.

16 . The at least one non-transitory computer readable medium of claim 15 ,
wherein,
the subset worker nodes are in an idle state,
the subset worker nodes each store the data; and
the assignment of the UDF includes a selection of at least one of the subset worker nodes to execute the UDF.

17 . The at least one non-transitory computer readable medium of claim 15 , wherein the subset of worker nodes are selected based on at least one of a distance to a location of the data, an available capacity, or an estimated time to complete current tasks.

18 . The at least one non-transitory computer readable medium of claim 15 , wherein the assignment of the UDF includes additional worker nodes of the plurality of worker nodes that have not stored the data.

19 . The at least one non-transitory computer readable medium of claim 15 , further comprising instructions, which when executed by the at least one processor causes the at least one processor to:
determine whether the UDF is eligible for parallel execution; in response to determination that the USF is eligible for parallel execution, generate a plurality of instructions to the subset of worker nodes, each instruction of the plurality of instructions to pre-fetch a portion of the data associated with the UDF; transmit the plurality of instructions to the subset of work nodes, wherein each instruction of the plurality of instructions is not transmitted to more than one node of the subset of worker nodes; and assign the UDF to the subset of worker nodes.

20 . The at least one non-transitory computer readable medium of claim 15 , wherein the assignment of the UDF to the subset of worker nodes is based on one or more parameters including completion time, degree of parallel computation desired, or power consumption.

Assignees

• Cisco Tech Inc

Inventors

• Shimamura Komei
• Saha Amit Kumar
• Dutta Debojyoti

Classifications

• G06F9/5027 Primary

  the resource being a machine, e.g. CPUs, Servers, Terminals · CPC title

• H04L67/5681 Primary

  Pre-fetching or pre-delivering data based on network characteristics · CPC title

• G06F9/5072

  Grid computing · CPC title

• G06F9/5044

  considering hardware capabilities · CPC title

• G06F9/5033

  considering data affinity · CPC title