So @Mark Takata (Adobe) and updates with regards to lambda?

Hey Paul! Nothing I can share unfortunately, but we are still continuing to explore options.

Has anyone tested this new service from AWS? https://aws.amazon.com/rds/aurora/babelfish/

Running SQL Server on Aurora PostgreSQL?

Yep

It's good, like your using SQL Server

Trying to publish a message to SNS using the Java SDK that has message attributes. Can anyone share an example?

We have started the upgrade from CF 2018 --> 2021. We want to migrate many sets of files from an on premise web application to AWS served through CloudFront and S3. During this process we will need to "register" in a DB the new path for the folder (AWS prefix). We don't register every file just top level directory. A typical upload will consist of ~2000-3000 small files (html, gif, png) and a few large files (+100 MB). My plan is to: 1. Mount the disk currently used for the Web Server to the CF server (both on-premise) read only 2. With CF web application pick the current registration (knows the current web server path) 3. Ask for the new folder location in the S3 bucket 4. Upload the files - I am guessing this will take considerable amount of time, longer than current CF page timeout 5. Update the registration to point to AWS S3 bucket With all of the S3 built-in functions as well as using AWS REST functions (currently using), what is the most efficient way to code the upload: • uploadDirectory - easiest to code, but don't know if it does any of parallel, multi-part or will be the most efficient loop through all files and upload individually: • parallelUploadFile • multipartUpload • uploadObject • uploadObject but use threading (arrayMap, structMap, traditional threading) • Using REST API (cfhttp) A couple other notes: We would be using our staging CF server, not our production server, to do this. We don't want to upload all files currently on the web server, we want to be selective. The directory structures on the web server are not an exact path match to AWS S3 buckets and therefore may need manual prompting. It is not just upload but we also need to update the DB. Thoughts/comments/suggestions?

If you're willing to live with a request timeout setting of more than 60 seconds, you should be able to upload a couple hundred MB of data in a single request. You could then use the uploadDirectory method of ColdFusion 2021's built-in S3 support. I don't personally know if uploadDirectory works in a multithreaded way — that would be a question for the Adobe engineers. The AWS Java SDK definitely is multithreaded, however. multipartUpload is really only good for very large files — a gigabyte or more in size. It doesn't sound like you have many of those.

I'm working on adding server-side encryption support to Ortus' S3SDK and I'm getting hung up on something that seems simple and I'm curious if anyone can help me here. Without getting bogged down in the details, I'm testing out customer-provided server-side encryption keys which are provided to the REST API via the HTTP headers documented here: https://docs.aws.amazon.com/AmazonS3/latest/userguide/ServerSideEncryptionCustomerKeys.html#specifying-s3-c-encryption • the encryption key needs to be a 256-bit base64-encoded, so I used

generatesecretkey( 'AES', '256' )

• The base64-coded MD5 hash of the key I tried two different ways-- ◦ by hashing the base64-encoded key

toBase64( hash( key ) )

◦ or hashing the original binary key

toBase64( hash( toBinary( key ) ) )

However, no matter what I do, I get this error back from the AWS REST API

The calculated MD5 hash of the key did not match the hash that was provided.

I'm having a hard time finding a specific reference online that shows examples of what an AES256 key and it's hash would look like, but I can see that my values are not matching examples I see in tickets online. Is anyone, perhaps more familiar with how AWS does stuff, able to see what I'm doing wrong here?

☝️ Got this working 😉

Lambda snap start was announced at re:Invent a couple of days ago. Seems to have a pretty significant impact on Java based apps. I wonder if this will breath any life into Fuseless. https://aws.amazon.com/blogs/aws/new-accelerate-your-lambda-functions-with-lambda-snapstart/