There are up-to-date and hands-on articles in this very wiki:

• Page Segmentation
• Text Recognition RNN
• Compute errors and confusions
• Working with Ground truth

Moreover, there are some other frequently asked questions collected in a FAQ section.

For a general overview you can look at some mind maps:

• OCRopus Workflow Diagram
• OCRopus Overview by Tomas Breuel (2014)

The project repository contains some IPython notebooks. Note that most of them are outdated (in the folder ./OLD).

• rnnlib vs ocropy
• IPython notebook on using ocropy, partly outdated but still useful.

There are also some articles on ocropy on the web:

• Extracting text from an image using Ocropus
• Training an Ocropus OCR model
• Homemade manuscript OCR (1): OCRopy
• How to train your own OCRopy models
• CIS OCR Workshop (how to OCR historical documents, with examples from ABBYY Finereader, Tesseract, OCRopy), in particular module 5 describes the full workflow with ocropus
• How to Digitize Texts with Open-Source Command-Line Optical Character Recognition (OCR) Software

You can ask questions about ocropy by opening issues, in the ocropy gitter chat on the mailing list.