All the 3 nodes are identical (same RAM, CPU, disks, …). This is managed by Atlas but they are all be in the same AWS region so lag should be minimal. All 3 nodes should be identical in data and schema including indexes. We don’t have special reporting replicas.
Need to check and get back on these → “What is the size of your data? Frequency of inserts? Size of documents inserted?”
oplog/hr shows in last 24 hours > 6GB/HR, almost close to 8GB/HR
cache activity on primary node I see like read into as 1.87GB/s and written from 206.488B/S
Different peaks for read/write IOPS. Peak read IOPS like 8.87K/s and write max is around 9.61K/S
