Viewing Instance Overview Data

1. Log in to the TaurusDB console.
2. Click in the upper left corner and select a region.
3. In the navigation pane, click Overview.
4. In the Instances by Status area, check the statuses of all TaurusDB instances under the current account.
   Figure 2 Checking instances by status
   

   Table 2 Status description

   Status	Description	Handling Suggestion
   Total instances	Total number of TaurusDB instances in all states	-
   Abnormal	Total number of TaurusDB instances in the Abnormal state	Submit a service ticket.
   Frozen	Total number of TaurusDB instances in the Frozen state	See Resource Freezing, Unfreezing, Release, Deletion, and Unsubscription.
   Pending reboot	Total number of TaurusDB instances in the Pending reboot state NOTE: Modifications to some parameters require an instance reboot before they can be applied.	Reboot instances.
   Available	Total number of TaurusDB instances in the Available state	-

Based on the configured alarm rules, you can view active alarms of all TaurusDB instances under the current account, including alarms in the Alarm (metric) and Triggered (event) states.

1. In the upper right corner of the Alarms area, click Create Alarm Rule to access the Cloud Eye console.
   • By default, the system has a built-in alarm rule, which can be modified, disabled, and deleted. For details, see Modifying an Alarm Rule.
   • Click Create Alarm Rule to create an alarm rule to monitor a metric or event for instances. For details, see Creating an Alarm Rule.
2. In the upper right corner of the Alarms area, select a time window and view alarm details.
   • The time window can be Last 1 hour, Last 6 hours, Last 12 hours, Last day, Last week, or Last month.

   • The Alarm Severity area displays the total number of alarms and the number of alarms of each severity. Alarm severities include Critical, Major, Minor, and Warning.
   • The Top 5 Instances by Total Number of Alarms area displays alarm statistics of the top 5 instances with the largest number of alarms. You can hover over an instance to view the number of alarms of each severity.
   • For details about critical alarms, see Table 3.

     Table 3 Critical alarm description

     Parameter	Description
     Instance Name	Name of the instance where the alarm was reported. After the page is refreshed, the latest alarm details will be displayed in real time.
     Status	You can view active alarms of all instances in the current region, including alarms in the Alarm (metric) and Triggered (event) states. Alarm: The metric value reached the alarm threshold, and an alarm has been triggered but not cleared for the resource. Triggered: An event configured in the alarm policy triggered an alarm.
     Alarm Type	Alarm type to which the alarm rule applies. Alarm (metric) Triggered (event)
     Alarm Policy	Policy for triggering an alarm. If you set Alarm Type to Metric, whether to trigger an alarm depends on whether the data in consecutive periods reaches the threshold. For example, Cloud Eye triggers an alarm if the average CPU usage of the monitored object is 80% or more for three consecutive 5-minute periods. For suggestions on handling high CPU usage, see What Should I Do If the CPU Usage of My TaurusDB Instance Is High? If you set Alarm Type to Event, the event that triggers the alarm is an instant operation. For example, if an instance fails to be created, an alarm is triggered. For details about supported events and handling suggestions for exceptions, see Events Supported by Event Monitoring.
     Alarm Rule	Name or ID of the alarm rule.
     Last Updated	Latest time when the alarm was triggered.
     Operation	Click Metrics. In the displayed dialog box, check the metric monitoring views in the selected time window.

The system performs diagnostics on instances by combining runtime data with intelligent algorithms and provides diagnosis results and optimization recommendations for detected anomalies.

Figure 3 Viewing anomalies

Click an abnormal diagnosis item to view the abnormal instances and related metric data. Table 4 provides the supported anomaly items and troubleshooting recommendations.

For example, when a high vCPU utilization anomaly is detected, click High vCPU utilization to view the affected instance information, CPU usage, and CPU usage trends. You can also click Details in the Operation column to view detailed diagnosis results, including node information, optimization suggestions, optimization cases, and metrics. Choose appropriate handling suggestions as required.

Figure 4 Diagnosis details

• Handling real-time sessions and corresponding SQL statements
  • Analyze root causes (OBT).

    Click Root Cause Analysis in the Operation column. The displayed dialog box shows details about blocked sessions and their root cause SQL statements.

  • Manually kill sessions.

    To end a single session, click Kill Session in the Operation column to restore the database. You can also select sessions and click Kill Session above the list to kill them in batches. You can select up to 50 sessions at a time.

  For details, see Managing Real-Time Sessions.

  Figure 5 Handling sessions
  

• Handling slow SQL queries
  • Export slow query log details.

    Click Export. In the displayed dialog box, select an export mode. Quick export and full export are supported. After the slow query log details are exported, click View All Export Records. In the displayed dialog box, click Download in the Operation column. After the download is complete, view the slow query log details in the ZIP package.

  • Throttle SQL statements.

    To add a SQL throttling rule, click SQL Throttling in the Operation column. In the displayed dialog box, specify SQL Type, Keyword, and Max. Concurrent Requests, and click OK. For details, see Configuring SQL Throttling.

  • Perform SQL diagnosis.

    Choose More > Diagnose in the Operation column. In the displayed dialog box, confirm slow SQL query information and click OK to start the diagnosis. In the displayed dialog box, view the SQL diagnosis results.

  • Check historical execution plans.

    Choose More > Historical Execution Plan in the Operation column to view the SQL execution plan summary.

  For details, see Viewing Slow Query Log Details.

  Figure 6 Handling slow SQL queries
  

• Increasing specifications

  Go to the page for changing instance specifications. For details, see Changing the vCPUs and Memory of a DB Instance or Node Manually.