Webodm + clusterodm + nodeodm on AWS

[juliano_duarte]

Hello guys,

I’m trying to start this configuration… webodm + clustered + nodeodm everything running on AWS.

So… I’m running webodm, clusterodm and locked nodeodm in the same virtual machine.

So far so good…

I have had configured the aws-config.json to launch the ec2 in my was account using the auto-scaling and it works!

Around 10 minutes after launch the virtual machine, I receive the message:

warn: Cannot create machine: Error: docker-machine exited with code 1
warn: Could not remove docker-machine, it’s likely that the machine was not created, but double-check!
warn: Cannot create node via autoscaling: Cannot create machine (attempted 1 times)
warn: Cannot forward task 4e366244-40de-4915-bcf7-1ca696768ac2 to processing node 192.168.10.20:3001: No nodes available (attempted to autoscale but failed). Try again later.

Please, someone could help me?

PS: I recorded all nodeodm before.

Best Regards!

[Saijin_Naib]

Do you know if your user in the VM has access to docker or root?

If you do id $USER, you should see docker on the list.

If not, you might need to do sudo usermod -aG docker $USER.

[juliano_duarte]

Hi man,

Yes, I can see docker in the output of the command.

I`m trying again… but I have the same problem…

Best

[James_Thurstan]

Hey Juliano,
Sorry to hijack your post but are you able to share your AWS config file? I am down a similar path to you and am having the exact same issue, however the fact that they’re showing up in your clusterODM UI means you’ve gotten further than me. Mine seems to load the instance fine with docker but simply doesn’t install the nodeODM image. I have also found a bunch of other resources so I am more than happy to share as they might make more sense to you.
Thanks and best of luck if you haven’t figured it out already!

[juliano_duarte]

Hello James,

Don’t worry man! Sorry about delay, but I was busy in my work and only today I’m returning to this project.

So… If you are using Amazon AMI 2023 I recommend you change to Amazon Linux 2. I did it, and works very well.

Please, share your current problem, and I will be happy to help you!

Following… this is my configuration file:

{
“provider”: “aws”,

"accessKey": "",
"secretKey": "",
"s3":{
    "endpoint": "s3.us-east-1.amazonaws.com",
    "bucket": ""
},
"vpc": "",
"subnet": "",
"securityGroup": "",

"monitoring": false,
"maxRuntime": -1,
"maxUploadTime": -1,
"region": "us-east-1",
"zone": "b",
"tags": ["type,clusterodm"],

"ami": "ami-01ca642ccb7532191",
"iamrole": "instance-profile",

"spot": false,
"imageSizeMapping": [
    {"maxImages": 40, "slug": "c5a.8xlarge", "spotPrice": 1.0, "storage": 100},
    {"maxImages": 80, "slug": "c5a.8xlarge", "spotPrice": 1.0, "storage": 100},
    {"maxImages": 250, "slug": "c5a.8xlarge", "spotPrice": 1.0, "storage": 100},
    {"maxImages": 500, "slug": "c5a.8xlarge", "spotPrice": 1.0, "storage": 320},
    {"maxImages": 1500, "slug": "c5a.8xlarge", "spotPrice": 1.0, "storage": 640},
    {"maxImages": 2500, "slug": "c5a.8xlarge", "spotPrice": 1.0, "storage": 1200},
    {"maxImages": 3500, "slug": "c5a.8xlarge", "spotPrice": 1.1, "storage": 2000},
    {"maxImages": 5000, "slug": "c5a.8xlarge", "spotPrice": 1.1, "storage": 2500}
],

"dockerImage": "opendronemap/nodeodm",
"usePrivateAddress": true

}

Now, I’m receiving the following message at the and of processing…

Compressing all.zip

Uploading 1 files to S3…
Cannot upload file to S3: null (InvalidArgument), retrying… 0
Cannot upload file to S3: null (InvalidArgument), retrying… 1

Please, someone could help me if this error? I’m trying to change the code directly but I do not know program in nodejs.

Best Regards

[James_Thurstan]

Hey Juliano,
Thanks so much for your help! I ended getting mine to work quite well after about a week of constant troubleshooting. Occasionally it hits me with the “ECONNREFUSED” error but I think its to do with my spot pricing or something.
Anyways I also came across your error and made a change or two to my aws file.

{
“provider”: “aws”,
“accessKey”: “”,
“secretKey”: “”,
“s3”:{
“endpoint”: “s3.ap-southeast-2.amazonaws.com”,
“bucket”: “droneprocessing”,
“acl”: “none”
},
“vpc”: “”,
“subnet”: “”,
“securityGroup”: “ODMSecurity”,
“usePrivateAddress”: true,
“monitoring”: false,
“maxRuntime”: -1,
“maxUploadTime”: -1,
“region”: “ap-southeast-2”,
“zone”: “b”,
“tags”: [“type,clusterodm”],
“ami”: “ami-08939177c401ce8f9”,
“engineInstallUrl”: “"https://releases.rancher.com/install-docker/19.03.9.sh\”",
“spot”: true,
“imageSizeMapping”: [
{“maxImages”: 40, “slug”: “t3a.small”, “spotPrice”: 0.02, “storage”: 60},
{“maxImages”: 80, “slug”: “t3a.2xlarge”, “spotPrice”: 0.22, “storage”: 100},
{“maxImages”: 250, “slug”: “m5.large”, “spotPrice”: 0.1, “storage”: 160},
{“maxImages”: 500, “slug”: “m5.xlarge”, “spotPrice”: 0.2, “storage”: 320},
{“maxImages”: 1500, “slug”: “m5.2xlarge”, “spotPrice”: 0.4, “storage”: 640},
{“maxImages”: 2500, “slug”: “r5.2xlarge”, “spotPrice”: 0.6, “storage”: 1200},
{“maxImages”: 3500, “slug”: “r5.4xlarge”, “spotPrice”: 1.1, “storage”: 2000},
{“maxImages”: 5000, “slug”: “r5.4xlarge”, “spotPrice”: 1.1, “storage”: 2500}
],
“addSwap”: 1,
“dockerImage”: “opendronemap/nodeodm”
}

I think it may be to do with the “ACL” part. Also check your permissions with the S3 bucket as I also had to set mine to public during testing as I couldn’t be bothered to sort out permissions as it was giving me problems.
Another thing you may find interesting is Droneyard. I have seen a few people mention it but I am unsure if it is as effective as using a cluster setup. If anyone does have experience let me know otherwise I might run it myself and see how it goes. I was mostly hesitant about it not splitting a job and instead running everything on one instance as I couldn’t find anything about how it manages a job. Overall it does seem like more of a singular system which would be nice.

[juliano_duarte]

Hello James,

How’re you?

Thanks for share this information with me! I setup the S3 bucket with public access and worked!

Now… I’ll study how to do it works with ACL configured.

Thank you very much for your answer and shared this solution with me.

Best Regards…