Add test to isolate InitialMmapSize delay without read transaction#1137
Conversation
| // InitialMmapSize values when there is NO concurrent read transaction. | ||
| // This test helps isolate whether delays in TestDB_Open_InitialMmapSize are caused | ||
| // by the large InitialMmapSize itself or by reader-writer interaction. | ||
| func TestDB_Open_InitialMmapSize_Windows_NoReader(t *testing.T) { |
There was a problem hiding this comment.
This test is very similar to TestDB_Open_InitialMmapSize, please just update that case to include table driven sub tests
There was a problem hiding this comment.
Let's leave it as it's for now, this case is already good enough to verify the thought #1136 (comment).
Once it's verified, we can merge it into TestDB_Open_InitialMmapSize.
There was a problem hiding this comment.
I retried several times, and couldn't reproduce the issue so far. Please merge this test case into TestDB_Open_InitialMmapSize, so that we can two sub test (table driven). thx
2a39804 to
6edccd0
Compare
ahrtr
left a comment
There was a problem hiding this comment.
LGTM with several minor comments, thx
TestDB_Open_InitialMmapSize_Windows_NoReader verifies whether large InitialMmapSize values (1GB) cause write commit delays even without concurrent read transactions. This isolates whether the delays observed in TestDB_Open_InitialMmapSize are caused by the large InitialMmapSize itself or by reader-writer interaction. Signed-off-by: Mustafa Elbehery <elbeherymustafa@gmail.com>
6edccd0 to
c7755a2
Compare
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: ahrtr, Elbehery The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
This PR adds
TestDB_Open_InitialMmapSize_Windows_NoReaderto isolate the root cause of write commit delays on Windows with largeInitialMmapSizevalues.Steps
InitialMmapSize = 1GBon all platformscc @ahrtr @ivanvc