Details
-
Bug
-
Status: Closed
-
Major - P3
-
Resolution: Fixed
-
1.17.0
-
None
-
None
Description
In libmongoc 1.17.0, calls to mongoc_client_pool_destroy take roughly 500ms, or the default value for minHeartbeatFrequencyMS. In 1.16.2 this was near instantaneous.
I'm able to observe lower shutdown times by setting MONGOC_TOPOLOGY_MIN_HEARTBEAT_FREQUENCY_MS to lower values, but the shutdown time is always at least what that value is set to.
This behavior seems buggy, though I imagine it could be an intended part of the streaming changes made for 4.4 servers like the other ticket I field.
On 1.17, the following prints about 0.5 seconds. On 1.16.2, it prints 0.00x seconds.
int main(int argc, char *argv[]) { |
const char *uri_string = "mongodb://localhost:27017/?retryReads=true"; |
mongoc_uri_t *uri;
|
bson_error_t error;
|
|
|
mongoc_init();
|
|
|
uri = mongoc_uri_new_with_error(uri_string, &error);
|
if (!uri) { |
fprintf(stderr, |
"failed to parse URI: %s\n" |
"error message: %s\n", |
uri_string, error.message);
|
return EXIT_FAILURE; |
}
|
|
|
mongoc_client_pool_t *pool = mongoc_client_pool_new(uri);
|
|
|
mongoc_client_t *client = mongoc_client_pool_pop(pool);
|
mongoc_database_t *database = mongoc_client_get_database(client, "test"); |
mongoc_collection_t *collection = mongoc_database_get_collection(database, "blah"); |
|
|
mongoc_database_command_simple(database, BCON_NEW ("ping", BCON_INT32(1)), NULL, NULL, NULL); |
mongoc_client_pool_push(pool, client);
|
|
|
struct timeval tv1, tv2; |
gettimeofday(&tv1, NULL);
|
|
|
mongoc_client_pool_destroy(pool);
|
|
|
gettimeofday(&tv2, NULL);
|
printf ("Total time = %f seconds\n", |
(double) (tv2.tv_usec - tv1.tv_usec) / 1000000 + |
(double) (tv2.tv_sec - tv1.tv_sec)); |
return 0; |
}
|
Attachments
Issue Links
- related to
-
SWIFT-1008 Reuse MongoClients as much as possible in the tests
-
- Backlog
-