Skip to content

Conversation

@igorsol
Copy link

@igorsol igorsol commented Nov 15, 2024

No description provided.

igorsol added 29 commits May 29, 2024 19:54
…CursorExtend

When something is called by internal client read concern must be
explicitly specified.
also split out 'initial_syncer_fcb' library
reconstruct prepared transactions happens in _tearDown_inlock
Fix of the "Fatal message" happened in OplogCapMaintainerThread.
Both switch to dummy location and back to dbpath location should be
under single global lock.
The thread calling getInitialSyncProgress keeps opCtx and locks trying
to acquire initial syncer's mutex. This leaded to deadlock due to fact
that storage changing thread waited for destruction of opCtx while
keeping initial syncer's mutex
Initial syncer's shutdown() function is called twice: first time when
shutdown thread has not created opCtx yet. This allows us to wait for
the end of storage change operation and to ensure that no storage change
will be running whne shutdown() is called second time (with non-NOOP
opCtx already created). This is done to avoid deadlock between shutdown
thread and storage change.
@igorsol igorsol marked this pull request as ready for review July 15, 2025 12:30
@igorsol igorsol requested a review from ktrushin July 15, 2025 12:30
Copy link

@ktrushin ktrushin left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@igorsol igorsol merged commit 9528f5e into v7.0 Jul 18, 2025
1 check passed
@igorsol igorsol deleted the fcbis-70 branch July 18, 2025 13:46
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants