Alerts in SAP HANA Database and Data Lake

The following instructions demonstrate a few examples of triggering alerts in an SAP HANA database.

Ensure these actions are performed on a non-production system such as a trial account or only use the test alert.

1. Open the SAP HANA database explorer.

   

2. Execute the following SQL in the SQL Console to trigger a high (indicated by the parameter value of 4) severity test alert.

   CALL _SYS_STATISTICS.Trigger_Test_Alert(?, 4, 'High test alert');  
   

   The alert will be viewed in SAP HANA Cockpit in step 3.

3. Execute the following SQL to trigger long-running statements (ID 39) at each threshold assuming the thresholds were set to 3, 2, and 1 minutes with an interval time set to 1 minute.

   DO BEGIN
     -- Wait for a few seconds
     USING SQLSCRIPT_SYNC AS SYNCLIB;
     CALL SYNCLIB:SLEEP_SECONDS( 300 );  --runs for 5 minutes
     -- Now execute a query
     SELECT * FROM M_TABLES;
   END;
   

   Long running statements can be run as a background activity which will continue even if the machine the SAP HANA database explorer is running on loses internet connectivity or the tab is closed.

   

   For additional details on the database explorer see the tutorial Get Started with the SAP HANA Database Explorer.

4. The alert record count of non-partitioned column-store tables (ID 17) can be triggered by executing the following SQL.

   --The default threshold for 'Record count of non-partitioned column-store tables' is 300 million
   -- This SQL may take a minute or two to run
   -- Create a table and insert more than 300 million rows into it
   DO BEGIN
          DECLARE i INT;
          CREATE TABLE MYTABLE(MYVALUE INT);
          INSERT INTO MYTABLE VALUES(1);
          INSERT INTO MYTABLE VALUES(2);
          INSERT INTO MYTABLE VALUES(3);
          INSERT INTO MYTABLE VALUES(4);
          INSERT INTO MYTABLE VALUES(5);
          FOR i IN 1 .. 26 DO
                 INSERT INTO MYTABLE (SELECT * FROM MYTABLE);
          END FOR;
          SELECT COUNT(*) FROM MYTABLE;
   END;
   SELECT TOP 100 * FROM MYTABLE;
   
   -- To resolve, partition the table
   -- ALTER TABLE MYTABLE PARTITION BY HASH(MYVALUE) PARTITIONS 5;
   
   -- Clean up
   -- DROP TABLE MYTABLE;
   

   As the default interval time for this check by default is set to 1 hour, it can be manually trigged by pressing the Check Now button in the alert definitions app within the SAP HANA cockpit.

   

   Note that two other alerts may also be triggered by the above SQL; table growth of non-partitioned column-store tables and record count of column-store table partitions.

5. The Database Overview in the SAP HANA cockpit will indicate if a medium or high alert was triggered. The refresh interval can be set to control how often the contents of the page are updated.

   

The following instructions will show how to view a triggered SAP HANA database alert in the alerts app in the SAP HANA cockpit.

1. After the long-running statement completes, open the alerts app to view the triggered alerts.

   

   Current alerts are alerts that were triggered by the last scheduled run of the alert definition. You may need to select Past from the Type drop down.

   ---

   Alerts can be sorted and grouped.

   

   ---

   Details about a SQL statement from a long-running statement alert can be found out with the following query.

   SELECT * FROM M_SQL_PLAN_CACHE WHERE STATEMENT_HASH='XXXXXXXXXXXXXXXXXXXXX';
   

   

2. The test alert will resolve itself after 5 minutes or can be removed (indicated by the parameter value of 0) from the Alerts tile by executing the following statement.

   CALL _SYS_STATISTICS.Trigger_Test_Alert(?, 0, 'Resolve test alert');
   

   Additional details on the test alert are available at SAP Note 3004477 - Usage of statistics server test alert (ID 999).

   Alerts can also be accessed via a REST API as shown at Accessing SAP HANA Cloud Alerts and Metrics using a REST API.

The following instructions show one example of triggering the data lake locked user event. The alert will be triggered when a user attempts to log in after the user account has been locked because an incorrect password was provided too many times.

1. In a SAP HANA database explorer that is connected to a data lake, execute the following SQL to create a login policy and a new user.

   CREATE LOGIN POLICY lp max_failed_login_attempts=3;
   GRANT CONNECT TO user2 IDENTIFIED BY 'Password2';
   GRANT SELECT ANY TABLE TO user2;
   GRANT SET ANY CUSTOMER PUBLIC OPTION to user2;
   ALTER USER user2 LOGIN POLICY lp;
   

2. Create a new connection to the data lake using user2 but with an incorrect password.

   First choose Properties for the existing connection and copy the host value.

   

   Create a new connection using the copied host, port 443, user2 and an incorrect password.

   

   After pressing OK an attempt will be made to connect to the data lake. After three failed attempts with the incorrect password, user2 will become locked. This can be seen in the SAP HANA cockpit.

   

   The alert can pushed to a channel such as a specified email address which will be shown in step 5 the first time an attempt is made to login to an already locked account.

3. Additional details about users can be seen by calling the procedure sa_get_user_status. The user can be unlocked using by resetting the login policy.

   CALL sa_get_user_status;
   ALTER USER user2 RESET LOGIN POLICY;
   --DROP USER user2;
   --DROP LOGIN POLICY lp;
   

   The tutorial Monitor a Standalone Data Lake in SAP HANA Cloud may also be of interest as it demonstrates the data lake IQ monitoring views.

The Business Technology Platform (BTP) includes a service called the Alert Notification Service (ANS) that provides a common way for other services or applications running in the BTP to send out notifications such as an email, a post to a Microsoft Teams or Slack channel, the creation of a ticket in ServiceNow, or a webhook to send events to any Internet REST endpoint. The SAP HANA Cloud database and data lake pass on events to (ANS) when an alert is triggered.

For an overview of ANS and information about the different service plans including free tier, see SAP Alert Notification Service for SAP BTP in the SAP Discovery center and watch the associated video SAP Alert Notification service for SAP BTP - Overview.

In this step, ANS will be configured to act on the incoming notifications by sending an email with the details of the alert. First, an instance of the alert notification service will be created. Then, two conditions will be created; one that matches notifications sent from an SAP HANA database and one for the data lake. An email action will also be created that describes who to send an email to and what content to include in the email when one of the conditions occurs. Finally, a subscription will be created that will use the two conditions and the action. Having the conditions and actions separate from the subscription enables them to be reused in multiple subscriptions.

1. Create an instance of the Alert Notification Service in the SAP BTP Cockpit.

   

   The Alert Notification Service must be in the same cloud foundry org and space as the SAP HANA Cloud instances that it will be receiving notifications from.

   ---

   If the Alert Notification service does not appear, it may be that the entitlement needs to be added to the subaccount. To do so, navigate to the subaccount, select Entitlements, Configure Entitlements, Add Service Plans, and Alert Notification.

   

2. Provide a name for the alert notification service instance and press the Create button.

   

3. Once the instance has been created, click on Manage Instance.

   

4. Create a condition for SAP HANA database alerts by selecting Conditions and then pressing Create.

   

5. Specify the condition details and press Create.

   

   Alternatively, a condition resource.resourceType set to equal hana-cloud-hdb could be used.

   ---

   Note that the Mandatory checkbox is left unchecked. This means that this condition can be ORed with one or more other non mandatory conditions.

   

6. Create another condition for data lake alerts. Specify the condition details and press Create.

   

7. Create an action by selecting Actions and the pressing Create.

   

8. Specify the type of action to be Email.

   

   See also Managing Actions for details on other available action types.

9. Provide a Name and then scroll to the additional properties section.

   

   In the Additional Properties section, provide values for Email Address, Subject Template and Payload Template.

   

    

   {severity} {resource.resourceType} {eventType} {tags.ans:status} occurred on {resource.resourceName}
   

   AlertID: {tags.hanaAlertRuleId}
   Body: {body}
   Recommended Action: {tags.ans:recommendedAction}
   Instance Details: {resource.tags.*}
   

   It is also possible to leave the subject and payload template fields empty. In this case, a default template will be used.

   Details on the available tags for each alert are provided at Built-In Events.

   

10. A confirmation token will be sent to the email address. Click on the provided link or copy that value and use it to confirm the action as shown in the next sub-step.

    

11. Click on email_action to open it.

    

    Press the Confirm Action button.

    

    Enter the confirmation token.

    

12. Create a subscription by selecting Subscriptions and pressing Create.

    

    Provide a Name and press the Create button.

    

13. Assign the conditions to the subscription by including the two previously created conditions and pressing the Assign button.

    

    Another example is to specify the instance ID of a specific production server to monitor and to only send an alert when the alert first occurs (create state).

    

14. Assign the action to the subscription by including the previously created action and pressing the Assign button.

    

15. Completed subscription.

    

16. Repeating the actions in step 2 and 4 should now send emails like the ones shown below.

    Test Alert

    

    Test Alert (default template)

    

    Notice that the ans:status is CLOSE. Events can have a status of CREATE, UPDATE or CLOSE.

    Long-Running Statement

    

    Record count of non-partitioned column-store tables

    

    Locked user

    

This step will briefly show an example of how to receive a notification in Slack.

1. In a browser open api.slack.com, click on Create a custom app, and then choose from scratch.

   Provide an app name, a workspace, and press Create App.

   

2. Choose to add Incoming Webhooks.

   

3. After activating incoming webhooks, click on Add New Webhook to Workspace.

   Specify the Slack channel that will be used to post to and click Allow.

   

   Copy the provided Webhook URL.

4. Similar to the step where an ANS action was created for email, create one for Slack in the SAP BTP Cockpit.

   

   Provide the webhook URL.

   

   An example of a payload template.

   [{
   	"type": "section",
   	"text": {
   		"type": "mrkdwn",
   		"text": "* {severity} {eventType} {tags.ans:status} on {resource.resourceName}*"
   	}
     	},
     	{
     		"type": "divider"
     	},
     	{
     		"type": "section",
     		"text": {
     			"type": "mrkdwn",
     			"text": ":pencil2: Subject: {subject}"
     		}
     	},
     	{
     		"type": "section",
     		"text": {
     			"type": "mrkdwn",
     			"text": ":scroll: Body: {body}"
     		}
     	},
     	{
     		"type": "section",
     		"text": {
     			"type": "mrkdwn",
     			"text": " :medical_symbol: Recommended Action: {tags.ans:recommendedAction}"
     		}
     	}]
   

5. Assign the action to a subscription.

   

6. After triggering an alert, the Slack channel will show the notification.

   

Congratulations! You have now configured alerts, triggered and viewed them via the SAP HANA cockpit and received them via email, Microsoft Teams, and Slack using the Alert Notification Service (ANS).