Scaling example
The following example serves as an orientation for the scaling scheme described below. It represents the standard use case.
Number of users | 20 |
Number of documents | 5,000 |
Number of pages | 500,000 |
Fix and dynamic scaling of all services possible.
We recommend dynamic scaling via Horizontal Pod Autoscaler (HPA), included in the Helm Chart, which scales with CPU and RAM usage.
Deployment | Number of replicas | Process Step(s) | |
Min. | Max. | ||
|---|---|---|---|
Redaction Gateway | 1 | 3 | 1-3 |
Redaction Service | 2 | 10 | ABE |
Redaction Priority Service | 2 | 10 | 2 |
PDFtron Redaction Service | 1 | 4 | ABE + 3 |
Redaction Report Service | 1 | 3 | 3 |
Persistence Service | 1 | 2 | 1-3 + ABE |
Search Service | 1 | 4 | |
Image Classification Service | 3 | 9 | ABE |
Entity Recognition Service | 3 | 9 | ABE |
CV Analysis Service | 1 | 3 | ABE |
OCR Service | 1 | 4 | 1 |
Tenant User Management Service | 1 | 2 | 1+3+ABE |
Layout Paser Service | 1 | 4 | 1 |
The process step diagram below provides an overview of the process steps and helps you understand which steps the different services are involved in.
 |
Process steps