|
Measures reported by OraRacSysWaitTest
Like the wait activity on stand-alone Oracle servers, administrators also need to observe the wait activity on Oracle servers that are part of a Real Application Cluster (RAC). This test connects to the global view on an Oracle server in an RAC, and pulls out information pertaining to the system-level cluster-related events that wait on the global cache or any other global resource on that Oracle server. Using this information, administrators determine the cluster events on which the database spends most of its time, and which current connections are responsible for the reported waits.
Note:
This test needs to be enabled only while monitoring the Oracle servers in an RAC environment.
| Measurement |
Description |
Measurement
Unit |
Interpretation |
| Total_waits |
The total number of times waits happened on this event since the last measurement period
|
Number
|
If a high number of waits are observed on a specific event, you can use the detailed diagnosis capability of the OraRacSessionWaitTest to figure out whether any current connections have contributed to the increase in waits.
|
| Total_waits_timedout |
The total number of waits on this event that timedout since the last measurement period
|
Number
|
A large number of timedout wait events is typically, undesirable. Use the Oracle-specific documentation to probe the cause of the timeout.
|
| Avg_time_waited |
The average duration for which the waits on this wait event persisted since the last measurement period
|
Secs
|
By comparing the value of this measure across all monitored wait events, you can determine where the database spends most of its time.
|
|