Application I am working on requires text contents to be extracted from various proprietary document formats like Microsoft word documents (doc, ppt, xls ), pdf and etc.

I am planning to implement a micro-service which takes document in proprietary format as input and returns extracted text as output.

This solution requires micro-service to exchange large amount of data per request (of the order 1 MB to 100 MB). Expectation is that microserive should be able to scale to 1000 requests per second.

W.r.t to this solution want to understand

  • Is it OK to transfer data at this rate over micro-service architecture?
  • Planning to use rest API's to transfer data. Is it a good option ?
  • 1
    A minimum of 1MB per request, 1000 requests per sec. That's 1 GB per second transfer rate you need (at least, the maximum is up to 100 GB/s). If you have multiple micro services, multiply this as each service needs to retransfer that same data. Even DDR4 memory can't maintain more than 25 GB/s! Are you sure your requirements are realistic?
    – Sjoerd
    Commented Dec 26, 2017 at 21:45

2 Answers 2


There are some important aspects you should consider first.


Let's imagine the 100 MB file is received by the service A which transfers it to service B, which, in turn, uses service C to do the actual parsing of the proprietary format.

The wrong approach would be for the services A and B to start sending the file to the underlying service only after they completely received the file from the client:

enter image description here

Instead, as soon as they start receiving the file, they should stream it to the underlying service.

enter image description here

This means that you're not waiting the time it takes to transfer 100 MB three times, but only one time, plus the latency...


Latency, on the other hand, cannot be avoided. Every intermediary service would still have to open the HTTP/HTTPS connection to the underlying service, before starting to transfer the file.

If your micro-services are located in the same data center, chances are the latency is a matter of a few milliseconds. If the services are hosted in different data centers, the latency may grow. With a high number of intermediaries, this can become a problem, and it will affect even small requests.

Possible DOS

When using the streaming technique, you should check that you don't open yourself to a possible DOS attack. The risk is that the intermediaries will keep the HTTP connection as long as the client is sending the file. The DOS attack would then consist of sending lots of files at a very low speed in order to exhaust the connections that the services are able to process.

  • And how you should stream the receiving file before it receives everything? using an PROXY? 🤔 Commented Apr 28, 2020 at 19:37

My suggestion to you is to start from something very simple and evolve. Try AWS Lambda + API Gateway for start. Very simple start, auto scaling up to 1000 concurrent executions per region. If you will need more concurrent executions you can think about multi region load balancing. Or try to enlarge quota by opening request. The question is pricing. can be expensive. Consider caching results if same documents used.

More complex arch :

  1. Upload document to s3. (Only after upload finished call Lambda with doc's name)
  2. Call Lambda to analyze the document in s3. (After, delete Doc, storing hash of the doc with result in cache)
  3. Return result.

This is more wise architecture from different perspectives.

  • Scales very good for multiple clients for s3 upload.
  • Latency Lambda <-> S3 much better
  • Simple to develop
  • pricing
  • 1
    And you can replace "S3" with any cloud computing provider, or use your own servers on a local network. Really the important bit in this answer is to put it some place, process it elsewhere and report back to the user when it is done. Commented Dec 10, 2018 at 17:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.