1

we have redash configured in docker 1.12.6-1.0.1.el6
OS oracle 6
image: redash/redash:8.0.0.b32245
image: redis:5.0-alpine
image: postgres:9.6-alpine
clickhouse-server 21.6.6.51-2

We have a problem with executing requests to Clickhouse.
The show tables or select * from session limit 5 queries work, but select * from session limit 10 does not work, we see an error in the log:

[PID:793][ERROR][ForkPoolWorker-775] Connection error to: http://our_IP:8123 (ReadTimeout).
Traceback (most recent call last):
File “/app/redash/query_runner/clickhouse.py”, line 152, in run_query
q = self._clickhouse_query(query)
File “/app/redash/query_runner/clickhouse.py”, line 113, in _clickhouse_query
result = self._send_query(query)
File “/app/redash/query_runner/clickhouse.py”, line 92, in _send_query
raise Exception(“Connection error to: {} {}.”.format(url, details))
Exception: Connection error to: http://our_IP:8123 (ReadTimeout).
[2021-08-10 17:14:20,393][PID:793][INFO][ForkPoolWorker-775] task_name=redash.tasks.execute_query task_id=318a2269-1a0c-438d-a203-14bce8baaf7e task=execute_query query_hash=18b0e4746278e9cfab5bb1c00765c574 data_length=None error=[Connection error to: http://our_IP:8123 (
ReadTimeout).]
[2021-08-10 17:14:20,394][PID:793][ERROR][ForkPoolWorker-775] Task redash.tasks.execute_query[318a2269-1a0c-438d-a203-14bce8baaf7e] raised unexpected: QueryExecutionError(u’Connection error to: http://our_IP:8123 (ReadTimeout).’,)
Traceback (most recent call last):
File “/usr/local/lib/python2.7/site-packages/celery/app/trace.py”, line 385, in trace_task
R = retval = fun(*args, **kwargs)
File “/app/redash/worker.py”, line 84, in call
return TaskBase.call(self, *args, **kwargs)
File “/usr/local/lib/python2.7/site-packages/celery/app/trace.py”, line 648, in protected_call
return self.run(*args, **kwargs)
File “/app/redash/tasks/queries.py”, line 436, in execute_query
scheduled_query).run()
File “/app/redash/tasks/queries.py”, line 382, in run
raise result
QueryExecutionError: Connection error to: http://our_IP:8123 (ReadTimeout).

1 Like
2

Looks like your clickhouse server terminated the connection and left Redash waiting. What do your clickhouse logs show?

3

Hello,
There are no errors in the clickhouse log.

4

it looks very strange, some queries are executed for example:
SELECT count(*)
FROM search

but the request will not work:
SELECT *
FROM search
limit 10

and with less limits it is executed successfully
SELECT *
FROM search
limit 5

I think some kind of clipboard is very small, to which the result is added before being output to redash

5

Yes it seems like a buffer somewhere isn’t configured properly. Just to be sure: do these queries work if you run them from a separate tool?

6

yes, all requests via clickhouse-client work from the same server where redash is configured

7

Sounds like a bug in our query runner, then.

Happy to review a pull request implementing a fix. Not something we can take on first-class since we don’t use Clickhouse or have a demo environment to test against.