Maximum number of uploads processed in parallel

  • Last Post 02 December 2013
fasterap posted this 24 November 2013


I just bought a package for 5K scans to test the performance of the API.

I am trying to process 500 images in parallel. The 500 have been queued but it seems that the API is only processing between 3 and 6 (status of "In Progress") at a given time instead of the 500 in parallel - The others just have the status of "Queued".

This means that processing my 500 images is taking a very long time where I was expecting the API to process them all in parallel.

Is there a way to process all of them in parallel to provide decent performance? Is there a limitation of the API and I should consider another solution?


Order By: Standard | Newest | Votes
Anastasia Galimova posted this 29 November 2013

The service increments its performance dynamically. We have some basic recourses, it allows to process single images immediately, and if the user sends a lot of images, they are queued. When the queue is big enough, the performance is automatically incremented by a fixed value. Recognition becomes faster in about a minute. The performance could be increased every 10-15 minutes.

We can discuss the details and help you with testing, if you send to the description of your project.

fasterap posted this 02 December 2013

Hi Anastasia, the performance definitely did not increase to acceptable levels. I upload 500 images together and only between 3 and 6 were every "in process" at the same time, all of the others being queue. I'll send a separate email to to get more info on how we could improve performance. Thanks!