Skip to content
This repository has been archived by the owner on Dec 15, 2021. It is now read-only.

Looks like a nightmare to get going with #36

Open
samuelgoldenbaum opened this issue Aug 18, 2017 · 7 comments
Open

Looks like a nightmare to get going with #36

samuelgoldenbaum opened this issue Aug 18, 2017 · 7 comments

Comments

@samuelgoldenbaum
Copy link

2 years old and not exactly straight forward, any alternatives?

@samuelgoldenbaum samuelgoldenbaum changed the title Looks like a nightmare of get going with Looks like a nightmare to get going with Aug 18, 2017
@marcosnils
Copy link

Sorry, yes. I'd recommend using AWS Lambda to process the dynamo streams and send it to a ES instead of using this old and unsupported logstash plugin.

@mholford
Copy link

Would be nice if Amazon updated their documentation. They still list this as the recommended approach for loading streaming data from Dynamo -> ES

@danbovey
Copy link

danbovey commented Feb 2, 2018

You're right, this is ridiculous. And I'm all the way down this rabbit hole, all because DynamoDB can't do a lowercase text contains query.

Edit - to add something helpful to this. The alternative tutorial on the AWS blog that uses lambda is outdated as the blueprint no longer exists. But looks like somebody has written a python script here: https://github.com/vladhoncharenko/aws-dynamodb-to-elasticsearch/blob/master/scripts/dynamodb-to-es.py

@aetter
Copy link

aetter commented May 2, 2018

Hi everyone, I recently removed the reference to this plugin from the Amazon Elasticsearch Service Developer Guide and added some simple sample code for stream processing:

https://docs.aws.amazon.com/elasticsearch-service/latest/developerguide/es-aws-integrations.html#es-aws-integrations-dynamodb-es

Very open to feedback. Feel free to open issues here:

https://github.com/awsdocs/amazon-elasticsearch-service-developer-guide/issues

@marcosnils
Copy link

+1 thanks

@scenterbar-sans
Copy link

+1 thanks

Did you follow through with the lambda solution? I feel like the solution in the docs doesn't account for failures or the inability to connect to es for any period of time. You would need to implement dead letter queues, and separate lambda functions for full table loads as opposed general streaming. @aetter if you have any recommendations I would certainly appreciate it. I just spoke with an elasticsearch subject matter expert from aws and his recommendation was to use logstash, which I agree seems much more robust. I would seriously appreciate any input or direction.

@mrichman
Copy link

I just spent 2 hours fighting with Java to get this working. Then I found this issue. I'm abandoning this effort. Someone should update the README to indicate this "solution" is hopelessly deprecated.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

7 participants