Computed pattern - More writes then reads, still viable


So I just read about the Computed Pattern and I had a question about it. Is it also worth it to implement this if your write count per hour is higher than the read count per hour?

Let me give you a little more background information about our project: we are receiving millions of messages per day from public transportation companies with information about cancellations, delays, vehicle positions and more. We are currently storing this data in an SQL database and we have built a dashboard to access that data. We also check if a received message from a public transport company matches to a timetable, if it doesnt match (called a mismatch) we count how many times that occurs for a given company and this is stored in a different table. But this data is only viewed once or twice per hour by our internal data management team. So the write count is way higher than view count.

We are investigating if MongoDB works for us instead of SQL, so is it still worh to implement this Computed Pattern for this “mismatch” entity, or is there another pattern better suited for this use case?

I would say it worths more. Writing is more work than reading. Indexes must be updated. Updated data mist be written to disk and replicated.

I’m sorry, I don’t understand your reply. Do you mean in this case still use the computed pattern? Even though the write count is higher than the read count?

yes, that is what i mean