[SERVER-30240] remove incorrect paragraph from getNextOpTime() comments in oplog.cpp - MongoDB Jira

XML

Word

Printable

JSON

Details

Type: Bug
Status: Closed
Priority: Trivial - P5
Resolution: Fixed
Affects Version/s: None
Fix Version/s: 3.5.11
Component/s: Replication
Labels:
None

Backwards Compatibility:
Fully Compatible
Operating System:
ALL

Description

The follow paragraph is no longer true and should be removed. We should link this ticket to the SERVER ticket in which this assumption changed.

https://github.com/mongodb/mongo/blob/878e55a460de02de2f0171d3e9e1bcfc12bb80cc/src/mongo/db/repl/oplog.cpp#L148


/**
 * Allocates an optime for a new entry in the oplog, and updates the replication coordinator to
 * reflect that new optime.  Returns the new optime and the correct value of the "h" field for
 * the new oplog entry.
 *
 * NOTE: From the time this function returns to the time that the new oplog entry is written
 * to the storage system, all errors must be considered fatal.  This is because the this
 * function registers the new optime with the storage system and the replication coordinator,
 * and provides no facility to revert those registrations on rollback.
 */
void getNextOpTime(OperationContext* opCtx,
                   Collection* oplog,
                   ReplicationCoordinator* replCoord,
                   ReplicationCoordinator::Mode replicationMode,
                   unsigned count,
                   OplogSlot* slotsOut) {

Attachments

Issue Links

related to

SERVER-15192 Make all logOp listeners rollback-safe

Closed

Activity

People

Assignee:: Andy Schwerin

Reporter:: Benety Goh

Participants:: Andy Schwerin, Benety Goh, Githook User

Votes:: 0 Vote for this issue

Watchers:: 2 Start watching this issue

Dates

Created:: Jul 20 2017 04:00:34 PM UTC

Updated:: Oct 12 2017 08:55:46 PM UTC

Resolved:: Jul 21 2017 12:44:48 PM UTC

oplog.cpp
143	/**
144	* Allocates an optime for a new entry in the oplog, and updates the replication coordinator to
145	* reflect that new optime. Returns the new optime and the correct value of the "h" field for
146	* the new oplog entry.
147	*
148	* NOTE: From the time this function returns to the time that the new oplog entry is written
149	* to the storage system, all errors must be considered fatal. This is because the this
150	* function registers the new optime with the storage system and the replication coordinator,
151	* and provides no facility to revert those registrations on rollback.
152	*/
153	void getNextOpTime(OperationContext* opCtx,
154	Collection* oplog,
155	ReplicationCoordinator* replCoord,
156	ReplicationCoordinator::Mode replicationMode,
157	unsigned count,
158	OplogSlot* slotsOut) {