It seems that Atlas checks if the relationship exists within the uploaded batch, rather than between the batch and already uploaded entities. There are a couple of approaches you could consider to avoid timeouts during bulk uploads:

Uploading Related Entities in Separate Batches: It’s possible to upload related entities in separate batches. However, you need to ensure that dependencies are respected between batches. If relationships between entities need to be established, you may have to upload them in an order that ensures the relationships can be checked and linked after the entities are uploaded.

Batch Size Management: If timeouts are an issue, you might want to consider reducing the batch size for uploads. Smaller batches can reduce the load on the system and help avoid timeouts. This might involve splitting larger datasets into smaller, more manageable chunks.

Optimize Atlas Configuration: Adjusting some configurations in Atlas, such as increasing the batch size limit or optimizing the database (e.g., using indexing) might help to handle larger uploads efficiently.

Asynchronous Upload Strategy: If possible, you can consider uploading entities asynchronously to prevent long-running operations that can lead to timeouts. This allows the system to handle multiple requests in parallel without overwhelming it.

Increase Timeout Settings: If you're encountering timeouts during bulk uploads, you could also look into adjusting timeout settings for the upload process, either at the Atlas server or API level, if that's a feasible option.

If you want to upload everything in one batch but avoid timeouts, breaking down the process into smaller, logical steps, while maintaining the required relationships, is usually the most effective approach.

question Unable to upload related entities separately in Support Questions

Unable to upload related entities separately

Re: Unable to upload related entities separately