[SERVER-82054] Investigate and fix slow bulkWrite request parsing on mongod Created: 10/Oct/23  Updated: 10/Nov/23  Resolved: 10/Nov/23

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

Type: Task Priority: Major - P3
Reporter: Vishnu Kaushik Assignee: Backlog - Replication Team
Resolution: Won't Do Votes: 0
Labels: milestone-1
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Depends
is depended on by SERVER-82230 Allow bulkWrite to accept array of do... Closed
Related
related to SERVER-82055 Investigate and fix slow bulkWrite re... Closed
related to SERVER-82056 Investigate and fix slow bulkWrite ad... Closed
is related to SERVER-76662 bulkWrite perf testing against vector... Closed
Assigned Teams:
Replication
Participants:

 Description   

As per the flame graph, currently about 4% of the bulkWrite operation thread's time is spent in parsing the bulkWrite request with 1000 insert ops. In contrast, about 1% of an insert operation thread's time is spent parsing a similar insert with 1000 ops.

The parsing code is auto-generated through IDL so it may be harder to optimize without changing the format of the bulkWrite request altogether.

See SERVER-76662 for flame graphs.



 Comments   
Comment by Sean Zimmerman [ 10/Nov/23 ]

Covering this under other tickets

Generated at Thu Feb 08 06:48:07 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.