Job scheduling based on job execution history

1. A system, comprising:

a plurality of computing resources of a multi-tenant provider network; and

one or more computing devices configured to implement a job scheduling system, wherein the job scheduling system is configured to:

receive a request to schedule a set of jobs, wherein the set of jobs is associated with a user account;

determine a resource allocation score based at least in part on whether or not previous jobs completed successfully according to a job execution history for the user account;

determine, based on the resource allocation score, a quantity of the set of jobs to be scheduled for concurrent execution at a future time, wherein the determined quantity of jobs to be scheduled for concurrent execution varies for different values of the resource allocation score;

schedule a first portion of the set of jobs for concurrent execution at the future time, wherein the first portion comprises the determined quantity of the set of jobs, and wherein execution of a remaining portion of the set of jobs is delayed;

reserve one or more of the computing resources of the multi-tenant provider network for the first portion of the set of jobs; and

initiate, at the scheduled time, the concurrent execution of the first portion of the set of jobs using the reserved one or more of the computing resources of the multi-tenant provider network.

2. The system as recited in claim 1 , wherein the job scheduling system is further configured to:

monitor the execution of the first portion of the set of jobs;

determine a modified resource allocation score, wherein the modified resource allocation score is determined based at least in part on the execution of the first portion of the set of jobs;

schedule a second portion of the set of jobs for concurrent execution, wherein a quantity of jobs in the second portion of the set of jobs is determined based at least in part on the modified resource allocation score; and

initiate the concurrent execution of the second portion of the set of jobs using the computing resources of the multi-tenant provider network.

3. The system as recited in claim 1 , wherein the set of jobs is associated with one or more job identifiers, and wherein the resource allocation score is determined based at least in part on a job execution history for the one or more job identifiers.

4. A computer-implemented method, comprising:

determining a resource allocation score based at least in part on whether or not previous jobs completed successfully according to a job execution history for a user account;

determine, based on the resource allocation score, a quantity of a set of jobs to be scheduled for execution at a future time, wherein the quantity of jobs to be scheduled for execution varies for different values of the resource allocation score;

scheduling a first portion of the set of jobs for execution at the future time, wherein the first portion comprises the determined quantity of the set of jobs, and wherein execution of a remaining portion of the set of jobs is delayed; and

initiating, at the scheduled time, the execution of the first portion of the set of jobs.

5. The method as recited in claim 4 , wherein the set of jobs is associated with the user account.

6. The method as recited in claim 4 , wherein the set of jobs is associated with one or more job identifiers, and wherein the resource allocation score is determined based at least in part on a job execution history for the one or more job identifiers.

7. The method as recited in claim 4 , wherein the set of jobs is associated with a project identifier, and wherein the resource allocation score is determined based at least in part on a job execution history for the project identifier.

8. The method as recited in claim 4 , further comprising:

monitoring the execution of the first portion of the set of jobs;

determining a modified resource allocation score, wherein the modified resource allocation score is determined based at least in part on the execution of the first portion of the set of jobs;

scheduling a second portion of the set of jobs for execution, wherein a quantity of jobs in the second portion of the set of jobs is determined based at least in part on the modified resource allocation score; and

initiating the execution of the second portion of the set of jobs.

9. The method as recited in claim 8 , wherein monitoring the execution of the first portion of the set of jobs comprises, for an individual job in the first portion in the set of jobs:

determining whether a launch of the individual job is successful or unsuccessful; and

if the launch is successful, determining whether an exit of the individual job is successful or unsuccessful.

10. The method as recited in claim 8 , wherein monitoring the execution of the first portion of the set of jobs comprises, for an individual job in the first portion in the set of jobs:

determining whether the individual job has continued to execute without an unsuccessful exit for a period of time.

11. The method as recited in claim 4 , further comprising:

increasing the resource allocation score based at least in part on successful aggregate execution of the first portion of the set of jobs; or

decreasing the resource allocation score based at least in part on unsuccessful execution of one or more jobs in the first portion of the set of jobs.

12. The method as recited in claim 4 , further comprising:

determining that the resource allocation score has fallen below a threshold score; and

suspending further execution of the set of jobs.

13. The method as recited in claim 4 , wherein the resource allocation score is determined using automated implementation of one or more policies by a job scheduling system.

14. The method as recited in claim 4 , further comprising:

reserving one or more computing resources of a multi-tenant provider network for the execution of the first portion of the set of jobs, wherein the first portion of the set of jobs is executed using the one or more computing resources.

15. The method as recited in claim 4 , further comprising:

reserving one or more computing resources of a compute environment for the execution of the first portion of the set of jobs, wherein one or more resource types of the computing resources are selected based at least in part on the resource allocation score.

16. A non-transitory computer-readable storage medium storing program instructions computer-executable to perform:

receiving a request to schedule a workload comprising a plurality of jobs, wherein the plurality of jobs is associated with a user account;

determining a resource allocation score based at least in part on whether or not previous jobs completed successfully according to a job execution history for the user account;

determining, based on the resource allocation score, a quantity of the plurality of jobs to be scheduled for concurrent execution at a future time, wherein the determined quantity of jobs to be scheduled for concurrent execution varies for different values of the resource allocation score;

scheduling a first portion of the plurality of jobs for concurrent execution at the future time, wherein the first portion comprises the determined quantity of the plurality of jobs, and wherein execution of a remaining portion of the plurality of jobs is delayed; and

initiating, at the scheduled time, the concurrent execution of the first portion of the plurality of jobs.

17. The non-transitory computer-readable storage medium as recited in claim 16 , wherein the workload is associated with a user identifier, and wherein the resource allocation score is determined based at least in part on a job execution history for the user identifier.

18. The non-transitory computer-readable storage medium as recited in claim 16 , wherein the workload is associated with one or more job identifiers, and wherein the resource allocation score is determined based at least in part on a job execution history for the one or more job identifiers.

19. The non-transitory computer-readable storage medium as recited in claim 16 , wherein the program instructions are further computer-executable to perform:

monitoring the execution of the first portion of the plurality of jobs;

determining a modified resource allocation score for the plurality of jobs, wherein the modified resource allocation score is determined based at least in part on the execution of the first portion of the plurality of jobs, and wherein the modified resource allocation score represents at least in part an estimated likelihood of successful execution of the remaining portion of the plurality of jobs;

scheduling a second portion of the plurality of jobs for concurrent execution, wherein a quantity of jobs in the second portion of the plurality of jobs is determined based at least in part on the modified resource allocation score; and

initiating the concurrent execution of the second portion of the plurality of jobs.

20. The non-transitory computer-readable storage medium as recited in claim 16 , wherein the program instructions are further computer-executable to perform:

increasing the resource allocation score based at least in part on successful aggregate execution of the first portion of the plurality of jobs; or

decreasing the resource allocation score based at least in part on unsuccessful execution of one or more jobs in the first portion of the plurality of jobs.