公开(公告)号：US10587529B1

公开(公告)日：2020-03-10

申请号：US16139478

申请日：2018-09-24

Applicant: Amazon Technologies, Inc.

Inventor： Shashanka Krishnaswamy

IPC: G06F9/50 ,  H04L12/917 ,  H04L29/08 ,  H04L12/725 ,  H04L12/26

Abstract: This disclosure describes techniques implemented partly by a service provider network for dynamically selecting between router groups to account for scaling up and/or scaling down of managed instances that are serving concurrent requests. Router groups define bandwidth allowances for routers that restrict the number of machine instances the routers are allowed to manage for individual user accounts. As concurrency of requests increases for a user account, rather than using more routers in a router group with a smaller bandwidth allowance, the techniques described herein include selecting a different router group for the user account that defines larger bandwidth allowances for user accounts. The service provider network may determine that the number of concurrent requests for a user account has passed a threshold for a router group, and select a new router group for routing new requests and managing containers of instances that are processing the existing concurrent requests.

公开(公告)号：US11093493B1

公开(公告)日：2021-08-17

申请号：US16147494

申请日：2018-09-28

Applicant: Amazon Technologies, Inc.

Inventor： Shashanka Krishnaswamy ,  Surya Prakash Dhoolam ,  Bradley Nathaniel Burkett

IPC: G06F7/00 ,  G06F17/30 ,  G06F16/2453 ,  G06F16/22 ,  G06F16/2458

Abstract: Various embodiments of dynamically switching between query and scan for optimizing table reads are described. A particular client of a shared table can use various criteria in determining whether to query for individual entries in a table, or whether to scan the entire table. Criteria can be, in some embodiments, whether there are fewer than a threshold number of other clients accessing the same portion of the table as the particular client, or the size of the portion accessed by the client. If there are fewer than a threshold number of other clients accessing the same portion of the table, and the size of the portion is below a size threshold, for example, then querying individual entries in the table might be appropriate in those circumstances. By querying instead of scanning, for example, clients can reduce the total number of I/O operations per second on the shared table.