This issue can't be edited
Because it belongs to an archived project. Jira admins can restore projects from the archive.
SPIKE: Pagination issue for GETting a thread for discussions forum
Description
Reason for Variance
Release Notes
Steps to Reproduce
Attachments
- 22 Oct 2015, 10:38 PM
- 22 Oct 2015, 07:18 PM
Lucidchart Diagrams
Activity
Christopher Lee June 13, 2016 at 8:47 PM
FYI @Verdan Mahmood
Nimisha Asthagiri December 17, 2015 at 3:08 PM
@khalidR Here's the Forums performance work we discussed at our standup this morning. The description has links to some background information.
Nimisha Asthagiri November 30, 2015 at 5:40 PM
SGTM. It can provide a concrete acceptance criteria for that work from the performance/user experience perspective.
LeslieG November 30, 2015 at 4:42 PM
@Nimisha Asthagiri @Scott Dunn I think we should move to TNL and assign to @Robert Raposa to begin discussions work. Does that sound good?
Christopher Lee November 30, 2015 at 4:35 PMEdited
Additionally, besides GETting a singlethread, GETting a thread list also has a slow operation (we think it's trying to count the comments) on the threads retrieved (in this case 20 threads with 3000-4000 posts each) which can take 10s+. https://courses.edx.org/courses/course-v1:IDBx+IDB6x+2015_T2/discussion/forum GETting a thread list is a common call when viewing threads.
Details
Details
Assignee
Reporter
Story Points
Priority
UnsetZendesk Support
Linked Tickets
Zendesk Support
It was found during the loadtests of the discussions API that the pagination of responses for threads was not working correctly as seen here: https://openedx.atlassian.net/wiki/display/MA/Initial+Loadtesting#InitialLoadtesting-pagination
Threads with more comments tend to have more requests.
Here is some investigative work done by @Brian Beggs as presented at the last hackathon: https://openedx.atlassian.net/wiki/display/TNL/Forums+Pagination