Note
For OceanBase Database V4.2.1, this parameter was introduced in OceanBase Database V4.2.1 BP7. For other versions of OceanBase Database V4.2.x, this parameter was introduced in OceanBase Database V4.2.5.
Description
log_storage_warning_trigger_percentage specifies the write performance threshold in percentage for triggering a log disk fault.
Limitations
Query the parameter
In the
systenant or a user tenant, you can execute theSHOW PARAMETERSstatement to query the parameter or query theGV$OB_PARAMETERSview for the parameter.Modify the parameter
You can modify the parameter only in the
systenant.
Attributes
Attribute |
Description |
|---|---|
| Type | Integer |
| Default value | 0 |
| Value range | [0, 50]. We recommend that you set this parameter to a value less than 10. |
| Modifiable | Yes. You can use the ALTER SYSTEM SET statement to modify the parameter. |
| Effective upon OBServer node restart | No. The setting takes effect immediately. |
Considerations
The log disk fault detection algorithm continuously collects baseline data about the write performance of the log disk, monitors the current write performance, and compares it with the baseline. If the current write performance is significantly lower than the baseline and does not recover within a period of time, the algorithm marks the log disk as faulty.
The relationship between the
log_storage_warning_trigger_percentageand log_storage_warning_tolerance_time parameters is as follows:- If the value of
log_storage_warning_trigger_percentageis0, no performance data is detected. Based on the original logic, the log disk is considered faulty when the write latency of the log disk is greater than the value oflog_storage_warning_tolerance_time. - If the value of
log_storage_warning_trigger_percentageis greater than 0, the log disk is considered faulty when its current write performance is lower than the value oflog_storage_warning_trigger_percentagemultiplied by the baseline and the duration of the situation reaches the value oflog_storage_warning_tolerance_time. log_storage_warning_trigger_percentagespecifies the write performance threshold in percentage for triggering a log disk fault, and a larger value indicates higher sensitivity of the fault detection algorithm.log_storage_warning_tolerance_timespecifies the maximum duration of performance degradation allowed, and a larger value indicates higher robustness of the fault detection algorithm.
- If the value of
Examples
Set the write performance threshold for triggering a log disk fault to 5% and the maximum duration of performance degradation allowed to 3s. If the write performance of the log disk is degraded by 5% compared with the baseline and does not recover within 3s, the log disk fault detection algorithm determines that the log disk is faulty.
obclient> ALTER SYSTEM SET log_storage_warning_trigger_percentage = 5;
obclient> ALTER SYSTEM SET log_storage_warning_tolerance_time = '3s';