[SERVER-54367] Improve the performance of asynchronous buildInfo Created: 05/Feb/21  Updated: 23/Mar/23  Resolved: 18/Mar/22

Status: Closed
Project: Core Server
Component/s: Internal Code
Affects Version/s: None
Fix Version/s: None

Type: Improvement Priority: Major - P3
Reporter: Amirsaman Memaripour Assignee: Backlog - Service Architecture
Resolution: Won't Do Votes: 0
Labels: re-triaged-ticket, servicearch-wfbf-sprint
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Depends
Related
Assigned Teams:
Service Arch
Participants:
Linked BF Score: 0

 Description   

The new, asynchronous implementation of buildInfo (see here) imposes some performance overhead on standalone in-memory benchmarks. The source of this overhead could be a mix of the following:

  • The context switch to schedule the rest of continuation on the buildInfo executor (see here).
  • The number of threads used to run asynchronous buildInfo commands (see here).
  • The cost of binding the client strand to the executor thread to serve buildInfo requests (see here). This is equivalent to the time it takes to construct the future-chain and releasing the client strand in here.

The acceptance criteria for this ticket is to reduce the cost of asynchronous buildInfo for the following benchmarks:



 Comments   
Comment by Lauren Lewis (Inactive) [ 18/Mar/22 ]

We haven’t heard back from you for some time, so we're going to close this ticket. If this is still an issue for you, please provide additional information and we will reopen the ticket.

Generated at Thu Feb 08 05:33:20 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.