| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
| |
Ensures that a collection lock is held in at least MODE_IS while
deregistering a PlanExecutor from the cursor manager. Introduces new
PlanExecutor::dispose() and ClientCursor::dispose() methods that must be
called before destruction of those classes, and ensures they are called
before destruction. These calls will thread an OperationContext all the
way through to DocumentSource::dispose() for each stage in a Pipeline,
which will give DocumentSourceCursor a chance to acquire locks and
deregister its PlanExecutor if necessary.
|
| |
|
| |
|
|
|
|
| |
This reverts commit 9e7974e4b6e2b3fe5e7741dce6549624113af196.
|
| |
|
|
|
|
|
|
| |
Moves registration of aggregation cursors to the global cursor manager.
This simplifies the logic for acquiring locks and resolving view
namespaces within the getMore and killCursors commands.
|
|
|
|
|
|
|
|
|
| |
Removes the class 'ScopedTransaction' and moves the responsibility of
abandoning the snapshot onto the GlobalLock class. Also renames the
AutoGetCollectionForRead class to AutoGetCollectionForReadCommand, and
adds a new AutoGetCollectionForRead class. Unlike
AutoGetCollectionForReadCommand, this new class will not update the
namespace on the CurrentOp object, nor will it add an entry to Top.
|
|
explain
Like other explainable commands, aggregate can now be
explained using the explain command, e.g.
db.runCommand({explain: {aggregate: ...}, verbosity:
"executionStats"}). The existing explain:true flag
corresponds to "queryPlanner" mode and is still supported.
However, explain:true cannot be specified when explaining
aggregate via the explain command.
Additional execution information is provided only in the
$cursor section of the aggregation explain output. Having
aggregation stages themselves track and report execution
info is further work.
|