eG Monitoring
 

Measures reported by TibTimeWrkldTest

Workload analysis for a Tibero database server involves:

  • Determining the number of transactions that applications execute on the database server at any given point in time;
  • Understanding the type of database operations these transactions trigger - executes? Updates? reads? Writes? Rollbacks? Parses?
  • Knowing how many users are active on the database server at a given point in time;
  • Determining how quickly the server processes this load and how much processing power was spent on the same.

This not only reveals the current workload of the database server, but also highlights the processing ability of the server, pinpoints bottlenecks in processing, and leads administrators to where these bottlenecks lie. To perform such detailed workload analysis, administrators can use the TibTimeWrkldTest test.

This test reports the current CPU usage of the server to indicate its current load. In addition, the test reveals the number and type of transactions the server processes every second, so that administrators can understand how well the server handles the load and can accurately identify where bottlenecks lie. By comparing the CPU usage of the server with its processing ability, administrators can intelligently figure out if the server requires additional CPU resources for improved performance.

Outputs of the test : One set of results for each database server being monitored.

The measures made by this test are as follows:

Measurement Description Measurement Unit Interpretation
Db_cpu_usage Indicates the percentage CPU used by the server. Percent A value close to 100% is indicative of excessive CPU usage. This in turn indicates that the server is using up all its processing power to service its current workload. It could be because the load is very high. It could also be owing to a few resource-intensive transactions executing on the server. In case of the former, you may want to allocate more CPU resources to the server, so as to enhance its processing ability.
Cpu_time Indicates the time for which the server has been hogging the CPU resources since the last measurement period. Secs A consistent increase in the value of this measure could indicate a steady increase in the workload of the server.
Redo_size Indicates the rate at which modifications were written to the redo logs since the last measurement period. MB/Sec If the value of this measure keeps growing, it could indicate that data is changing rapidly in the databases. A steady drop in this value could indicate that changes are not written to the redo logs as quickly as they occur.
Logical_reads Indicates the rate at which logical reads were performed by the server. Reads/Sec These measures are good indicators of the level of activity on the database server and how well the server handles these activity levels. In the event of a slowdown, you can compare the value of these measures to know where the slowdown may have originated - when making changes to data? When reading?
Block_chgs_count Indicates the rate at which database blocks were changed. Blocks/Sec
Phy_reads Indicates the rate at which the server performed physical reads. Reads/Sec
User_calls Indicates the rate at which the server made user calls. Calls/Sec  
Parses_count Indicates the rate at which the server parsed SQL statements. Parses/Sec Parsing is one stage in the processing of a SQL statement. When an application issues a SQL statement, the application makes a parse call to Tibero Database. During the parse call, Tibero Database:

  • Checks the statement for syntactic and semantic validity.
  • Determines whether the process issuing the statement has privileges to run it.
  • Allocates a private SQL area for the statement.

If the value of this measure keeps increasing consistently, it could indicate that many SQL statements are being executed on the server, thus generating more parses every second. If the value of this measure drops consistently, it could indicate a bottleneck in parsing.
Hard_parses_count Indicates the rate at which the server hard parsed SQL statements. Parses/Sec As opposed to a soft parse, a hard parse loads the SQL source code into RAM for parsing. If the value of this measure is decreasing steadily, it could mean that hard parsing is taking too long. It could also mean that very few hard parses are actually performed.
Logons Indicates the rate at which users login to the database server. Logons/Sec A steady rise in this value is indicative of a steady increase in user activity on the server.
Executes Indicates the rate at which executions are performed by the server. Executions/Sec  
Trans_rollbacks Indicates the rate at which the server performs rollbacks. Rollbacks/Sec Ideally, the value of this measure should be low. This is because, rollbacks are expensive operations and should be avoided at all costs. A consistent increase in the value of this measure is hence a cause for concern.
Tot_transactions Indicates the rate at which transactions were executed by the server. Trans/Sec A steady increase in the value of this measure could indicate an increase in the transaction load on the server. A consistent and notable drop in the value of this measure could indicate a bottleneck in transaction processing.