mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-11-10 09:32:06 +00:00
Merge pull request #29390 from azat/fix-async-insert-tests
Fix 02015_async_inserts_2 flakiness
This commit is contained in:
commit
369821634b
@ -106,13 +106,10 @@ private:
|
||||
QueryIdToEntry currently_processing_queries;
|
||||
|
||||
/// Logic and events behind queue are as follows:
|
||||
/// - reset_timeout: if queue is empty for some time, then we delete the queue and free all associated resources, e.g. tables.
|
||||
/// - busy_timeout: if queue is active for too long and there are a lot of rapid inserts, then we dump the data, so it doesn't
|
||||
/// grow for a long period of time and users will be able to select new data in deterministic manner.
|
||||
/// - stale_timeout: if queue is stale for too long, then we dump the data too, so that users will be able to select the last
|
||||
/// piece of inserted data.
|
||||
/// - access_timeout: also we have to check if user still has access to the tables periodically, and if the access is lost, then
|
||||
/// we dump pending data and delete queue immediately.
|
||||
/// - max_data_size: if the maximum size of data is reached, then again we dump the data.
|
||||
|
||||
const size_t max_data_size; /// in bytes
|
||||
|
@ -8,6 +8,9 @@
|
||||
<insert_quorum_timeout>60000</insert_quorum_timeout>
|
||||
<!-- 30 seconds (default is 2 minutes) -->
|
||||
<replication_wait_for_inactive_replica_timeout>30</replication_wait_for_inactive_replica_timeout>
|
||||
|
||||
<!-- Increase timeout for async INSERT to avoid flakiness in tests -->
|
||||
<async_insert_busy_timeout_ms>5000</async_insert_busy_timeout_ms>
|
||||
</default>
|
||||
</profiles>
|
||||
</yandex>
|
||||
|
Loading…
Reference in New Issue
Block a user