If your production is taking longer than expected, several factors can impact the processing speed in GoldFynch, especially when working with large or complex datasets. Below are the most common reasons for delays and how they might apply to your case.
1. High item count (over 100,000 files)
GoldFynch is optimized for speed and reliability, however when the item count exceeds 100,000 files, performance can slow down significantly. Large volumes require more system resources and time to process, tag, convert, and package into a production-ready format.
Tip: If you're approaching or exceeding this threshold, consider splitting the production into smaller, more manageable chunks. The system will process these smaller production sets in parallel, thus speeding up the production process.
Click here for information on splitting your production. (but see #4 before you go too small).
2. High page count
Even with a moderate number of files, if those files are lengthy or scanned PDFs with many pages, it can bog down the production. Each page needs to be rendered, processed, and stamped (if applicable), and that takes time, especially when producing to image formats like PDF/TIFF.
Tip: Check the page count in your production preview. A high page count will extend production time regardless of item count.
3. Redactions
When there are redactions in the production, the system must recreate every affected page, and that page must undergo the OCR process again, resulting in additional processing. Therefore, if your production includes redacted documents, expect extra time for the system to securely convert those pages into image versions.
Tip: The more redacted pages you have, the longer your production will take. Redactions are applied with precision, but that precision comes at the cost of time. Ensure that you completely delete any redactions you don't intend to maintain on a page. If there is even one redaction dot, the entire document is stripped of text and run through the OCR process again.
4. Too many small productions (e.g., 20K splits)
It might seem efficient to break up your production into smaller batches, such as 20,000-item chunks, but that can actually backfire. Each production batch carries overhead: system setup, metadata handling, hashing, file packaging, etc. Running 5 batches of 20K will take considerably longer than 2 batches of 50K.
Tip: When possible, aim for fewer, larger batches rather than many small ones. The system handles bigger batches more efficiently than multiple small ones processed sequentially.
Still Waiting?
If your production has been running for an extended period and none of the above applies, feel free to reach out to support at support@goldfynch.com. We're happy to check the backend and confirm everything is running smoothly.