Tracing Platform Data Use
In the SmartSpace release 3.2 there is a new tes_profile trace stream which is always enabled. Every five minutes, each running schema server prints a line to this trace stream. This profile stream can be used to detect issues with the performance of the system, such as spikes in load or continually growing table sizes. Eventually we intend to write an IT support feature which will record these traces and support reporting. For higher level methods of monitoring all aspects of system health, see Health monitoring.
The columns are separated by “|” and area as follows:
Column | Example | Meaning |
---|---|---|
Name of schema | Aggregation%Aggregated | The name of the schema as advertised to clients |
Transaction rate | 41.654 | The average number of transactions per second for this schema across the last five minutes |
Total row count | 403 | The total number of rows in all tables in the schema now |
Max row count | 403 | The number of rows in the largest table in the schema now |
Max row table | 0 | The index of the largest table within the schema now |