mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-12-05 05:52:05 +00:00
4669951db5
Reset the timeout for hedged requests after each packet (such as
Progress packets) to avoid triggering send_timeout/receive_timeout as a
total timeout for the query in some cases.
Here is a simple query that demonstrates the problem:
select * from remote('127.2', view(select sleep(3) from system.one)) settings receive_timeout=1
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
4 lines
477 B
SQL
4 lines
477 B
SQL
select * from remote('127.2', view(select sleep(3) from system.one)) settings receive_timeout=1, async_socket_for_remote=0, use_hedged_requests=1 format Null;
|
|
select * from remote('127.2', view(select sleep(3) from system.one)) settings receive_timeout=1, async_socket_for_remote=1, use_hedged_requests=0 format Null;
|
|
select * from remote('127.2', view(select sleep(3) from system.one)) settings receive_timeout=1, async_socket_for_remote=0, use_hedged_requests=0 format Null;
|