Error start nodes in Dremio AWS Edition

[l-jhon]

When starting the Dremio AWS Edition nodes on Elastic Engines, the following error is displayed:

Execution nodes didn't start within the start timeout.

Has anyone experienced this problem?

Thanks!!

[steven]

Usually when this happens, it is because either the IAM policy or Security Group is not configured correctly. Please make sure that you have configured them according to our documentation:

https://docs.dremio.com/deployment/aws/aws-edition-permissions.html
https://docs.dremio.com/deployment/aws/aws-edition-security-group.html

If you have done so, and still see this problem, we will need to do further investigation.

[l-jhon]

Thanks for the answer @steven

I’ve already done this validation, and the strange thing is that the nodes are created on AWS, I checked the EC2 Instance Dash and the nodes are there, but when the error message is displayed, they are finalized, this may be related to IAM policy or Security Group even when nodes are created on AWS?

[steven]

Even if the EC2 instances are created, if the IAM policy or Security Group prevents certain actions from happening during the initialization of the Dremio process on those instances, it will prevent them from properly joining the Dremio cluster

[l-jhon]

It was really a Security Group rule, lack of attention on my part!

Thanks for your help @steven

[Kevin_D]

Hi Jhon, we want to make sure that the instructions for setting up the AWS Edition are clear and easy to follow. Were the instructions for the Security Group setup confusing? Do you have any suggestions on making them easier to follow? Thanks in advance!

[l-jhon]

Actually the error occurred due to lack of attention on my part, but I believe that a change on the page https://www.dremio.com/tutorials/aws-step-by-step/ in the Configure Security Group part about what should be releases, help a little who is doing the deploy, but this is just an observation, the documentation is very good.

Thank you very much for your feedback and concern for the community.

[Kevin_D]

Thanks for the feedback Jhon. Did you use this tutorial to create your Security Group?https://www.dremio.com/tutorials/security-group-for-dremio-aws/

[l-jhon]

No, actually I used a Security Group that I had used before to manually deploy to an EC2 where I climbed a Standalone cluster, this Security Group had the rules for 9047, 31010, 22 and 443, it was just missing for All traffic self reference to SG. But now everything is fine, I will pay more attention to the documentation.

Thanks.

[mendeel]

Hello, guys!

I’m using AWS edition 11.00 with both SG and IAM Role setup automatically by the CloudFormation project that’s launched through the AWS Marketplace.

I’m having this exact problem as the OP. I can make the engines start, Dremio launches the EC2 instances that become “2/2 checks” but after some minutes, Dremio tells me “Execution nodes didn’t start within the start timeout.”.

I already checked the SG and changed the rule that was selfreferencing the SG “All TCP” to “All Traffic”. I even gave AdministratorAccess to the IAM Role but nothing helped.

Do you guys have any idea how to solve this?

Thanks!

[balaji.ramaswamy]

@mendeel Have you checked @l-jhon’s response on what solved the issue?

this Security Group had the rules for 9047, 31010, 22 and 443, it was just missing for All traffic self reference to SG. But now everything is fine

[mendeel]

Yeah, I was actually using the SG that was created by CloudFormation. I ended up creating a manual one by following the instructions on Dremio’s docs and it started working again

Thanks for your attention!