Hi, While generating elementary reports via

edr report

command with both elementary and edr version 0.16.1, my container (AWS ECS) exits with status code 137 (Out-of-memory) error. I wonder why generating report causes this error. Does

edr

fetches all failed columns or do another memory-intensive job?

Hey @Yunus Emre Gündüz! What DWH and orchestrator are you using?

I use Snowflake and Prefect.

We're aware of this issue and it's still in investigation on our side, we hope to have a solution for this by next week. For now, can you please share the edr logs (by DM) to help us investigate, and maybe try updating the Elementary CLI version? We'll update when we can!

I will try upgrading, thanks!

Hello, I also face this issue on Redshift + ECS, my container breaks with OutOfMemory error, I tried to clear my elementary schema but it did not work. We are running Elementary 0.16.1 Here are the error logs :

subprocess.CalledProcessError: Command '['edr', 'send-report', '-t', 'main', '--s3-bucket-name', 'sezane-preprod-eu-west-3-dbt', '--bucket-file-path', 'elementary/edr_report.html', '--profiles-dir', '.']' died with <Signals.SIGKILL: 9>.

And attached the memory utilization of the hourly container that run dbt + elementary

Hi, since our issue was happening randomly and seldomly, we are waiting for a solution.

Upgrading elementary solved the problem. In fac the command takes no memory.

Hearing that is so nice! Thank you 🙏

Hey @Félicien Barluet! Thank you for sharing. What Elementary version worked for you? 🙂

version = "0.18.1"

I prepared the environment for 0.18.1 but upgrade waits for the next cycle as it is not an hotfix due to rare & random occurrence of the error. I'll inform when the fix is deployed.

Thank you for updating! Great to hear.

Unfortunately, I recently upgraded a dbt project with lots of tests and it raised OOM error even after the upgrade from 0.16.1 to 0.18.1. It seems that the problem is not resolved.

@Yunus Emre Gündüz were you able to find a solution to this issue? I’ve been running into the same problem and am hoping to find a fix that doesn’t involve increasing ECS memory resources.