Feature

Size of the data that gets stored in Germain datastore depends on the 1) technology you are going to monitor 2) database technology you are going to use to store that data. Some guidelines and we can help with that sizing exercise:

Germain DB Sizing Example

Example of the Database Storage requirements for an environment that has 1500 concurrent users using a CRM and an ERP application, 8 hours per day. For this example, Germain retains 60 days of raw data, and 1 year of aggregated data.

Database Storage Estimate
Excluding Session ReplayAssumed Unit Size2048Bytes
Units/Day 1,000,000
Data retention - Raw/RCA Data60Day(s)
Data retention - Aggregated Data1Year(s)
Estimated Storage0.11TB
Session replay1-hour Session Replay Size31.00MB/hour
Concurrent users using your app1,500Concurrent users
Hours/day a user uses your app8Hour(s)/day
Data retention7Day(s)
Estimated Storage2.48TB

Total Estimated Storage2.60TB

Notes: If we use AWS, The 2000 IOPS is an option that AWS provides where they will guarantee the performance of the storage. We can always remove the "provisioned IO" and go to "GP2 (SSDs)". That would save some cost and the performance should still be good (on AWS large drives have faster IO in general).

Example: 2 ApplicationsMB/h
30s seek
interval
2 min seek interval5 min seek interval
ERP applicationExtreme User Activity31.0027.5026.80
Expected User Activity6.402.902.20
Idle User Activity4.001.200.47
CRM ApplicationExtreme User Activity10.407.506.90
Expected User Activity8.205.304.70
Idle User Activity3.901.000.39

Note:

  • The network traffic difference between two versions of a monitored application may significantly vary, so it is critical to perform a proper sizing analysis for each monitored app and each significant release.