I have a Community Health Workers (CHW) survey that I would like to implement using DHIS2.
Over a period of two years, I expect to collect around 2–3 million survey submissions, which means the Tracker database will contain millions of records.
Would DHIS2 Tracker be suitable for handling a dataset of this size? Has anyone implemented a Tracker program with a similar volume of data? I would also appreciate any recommendations regarding performance, scalability, or best practices for managing such a large number of records.
Without a doubt yes, you can handle millions of records in DHIS2 tracker if it is designed and configured properly.
In our projects, we have implemented DHIS2 Tracker for community health programs with large-scale data collection. Based on our experience, there are two important things to consider:
Use DHIS2 Tracker as the operational database, not as a long-term data warehouse. If you need consolidated reporting across multiple implementations, consider extracting the data to a centralized FHIR server or another data warehouse. We have multiple DHIS2 instances where data is kept separate, and common reporting can be achieved through plugins or integrations with a centralized FHIR server.
Design programs, program stages, program rules, and tracked entity attributes as efficiently as possible. Avoid unnecessary complexity, as it has a significant impact on performance.
Thank you very much, Mr. @RCPoudel, for your prompt response and for sharing your valuable experience. I truly appreciate your insights and the practical recommendations you provided.
I have one additional question, if I may.
Would it be a good practice to use DHIS2 primarily as a data collection platform only, while performing advanced analysis, reporting, and visualization in external tools such as Power BI or another data warehouse? In other words, is it advisable to treat DHIS2 mainly as an operational data collection system rather than the primary analytics platform?
Yes, both approaches are valid, and the right choice depends on scale, complexity, governance, and reporting requirements.
Here is a use case demonstrating Power BI for data visualization link
All data elements, indicators, and organizational units are fetched from the DHIS2 API.