Teradata administrators worldwide face similar challenges.
I propose creating a worldwide database containing the characteristics of Teradata systems. This will allow for easy comparison and analysis of performance issues between different systems.
My initial report displays the CPU’s idle time and the number of disks I/O waits for every hour throughout the day.
To contribute to the database construction, please execute the following SQL statement on your system:[su_panel border=”2px solid #cccccc” shadow=”8px 1px 2px #eeeeee”] SELECT
EXTRACT(HOUR FROM TheTime) time_of_day,
((sum(CPUIoWait)) / (sum(CPUIoWait+CPUUExec+CPUUServ+CPUIdle)))*100 as CPU_WAITING,
((sum(CPUIdle)) / (sum(CPUIoWait+CPUUExec+CPUUServ+CPUIdle)) ) *100 as CPU_IDLE
WHERE TheDate between DATE-30 AND DATE-1 GROUP BY time_of_day ORDER BY 1;[/su_panel]
The query will execute in just a few seconds and retrieve 24 records, each representing an hour.
Typically, the data is within the DBC.ResUsageSpma table is regularly purged and saved in a backup table. It is presumed that you have knowledge of the implementation and location of this process on your system.
Please provide details regarding the system type, number of nodes, and AMPs.
Please send your results to [email protected] (plain text or Excel Sheet).
I will collect all results and make them available online (charts & tables). Here is our current collection:
Kindly inform me of any additional measures that may interest you. I believe that a database of measures would be advantageous for everyone.