If I use Kinesis Firehose, that writes to let's say S3, as a source for Data Analytics, would it still be possible to store the original data from the stream, as if there were no Data Analytics integrated after Firehose? Or the original data, that arrived to the stream, will be lost and everything I'm gonna have will be output from Analytics step?
1 Answers
2
votes
The data injested by firehose to data analytics will also be written to your firehose destination directory parallely. So any data going to analytics is also being written to s3. This becomes your backup data. The data going out from data analytics will be fed to either firehose or stream or lambda whichever you are doing, from there it goes to another destination.
So to your question, there will be raw data as well as transformed data in s3 in seperate directories, as long as destination for first firehose and destination for analytics is different.
