redash_adhoc_worker eats CPU massively

[koooge]

Issue Summary

Hi there. redash_adhoc_worker sometimes becomes CPU eater and redash cannot work normally. When it becomes so, I reboot the machine to recover it, but it would not smart. So, I suppose redash needs to handle something adequately. Do you have any ideas?

$ docker stats

CONTAINER ID        NAME                        CPU %               MEM USAGE / LIMIT     MEM %               NET I/O             BLOCK I/O           PIDS
...
9ab7f894ad7d        redash_adhoc_worker_1       144.48%             400.8MiB / 3.794GiB   10.31%              4.11GB / 2.6GB      370MB / 147kB       3
...

Steps to Reproduce

1. This is the first step
2. This is the second step, etc.

Any other info e.g. Why do you consider this to be a bug? What did you expect to happen instead?

Technical details:

• Redash Version: 7.0.0+b18042
• Browser/OS: chrome
• How did you install Redash: EC2 (t3.medium) with AMI

[arikfr]

It's probably enough to restart the Docker container instead of the whole machine.

But it's worth understanding what's going on. What types of data sources are you using? How often does this happen?

[koooge]

rds_mysql, dynamodb or python (REDASH_ADDITIONAL_QUERY_RUNNERS=redash.query_runner.python). Actually I don't know what query does it cause because some users use this redash 😩

How often does this happen?

maybe 1 per 2week~month

[ranbena]

@koooge try this out #3903.
Waiting to hear that it works before we land it.

[koooge]

Thanks @ranbena ! It's not easy to ensure not to occur this. I close this issue since #3903 was merged. I will reopen this if it reproduces (it would be redash 8).