Alerts
DBA Dash doesn’t have a built in alert feature but it’s possible to construct custom alerts based on the data collected. Some examples are provided below. SQL Server also has some built-in alerting capabilities you utilize. DBA Dash can be used to compliment this.
A good alert system should provide fast notifications when issues are detected, allowing you to be responsive to incidents. It’s also important to limit the number of notifications generated. If your alert system generates alerts too frequently, alert fatigue will set in and any value in your alert system is lost. It’s better to be conservative about the alerts you configure and send notifications when the alert conditions provide a very strong indication that there is an issue that requires your attention.
SQL Server Built-in Alerts
SQL Server has some built in alerting capabilities that you might want to take advantage of. For example, SQL Server agent jobs can send a notifications on job failure, success or completion. You can cut down on the number of alerts generated by configuring the retry attempts and retry interval for your jobs. It’s also possible to configure alerts for corruption, high severity errors or performance condition alerts. Alerts can be configured with delays between responses to help cut down on the volume of alerts that might be received. I would reccommend alerts for error numbers 823, 824 and 825 and for severity 17-25.
Alerts provide immediate notification that there is a problem that requires your attention. This notification allows you to take fast corrective action that minimizes the impact of the issue. Alerts are great and you should use them - but a lack of notifications doesn’t mean everything is in a good state.
Some things to consider:
- What happens if someone disables my backup job?
- What happens if the SQL agent isn’t running?
- There is a problem with my DB mail setup?
- The alert got caught by a spam filter or email rule?
- You got a flood of alerts and the important alert got lost in the middle of that?
This is where a daily checks process can compliment your alerting system - providing validation that everything is in a good state. This is the purpose of the “Summary” tab in DBA Dash. Here you can see failed jobs, low disk space, missing backups, broken HA/DR and more. If you have alerts configured on your SQL instance it will also highlight when alerts were last triggered. This is a manual visual inspection - but the process is quick with the data you need on a single page. It’s designed to be used in combination with alerting.
Creating alerts from DBA Dash
DBA Dash collects the data from your SQL instances in a central repository. Although DBA Dash doesn’t have any built in alerting capabilities, you can build your own alerts based on the data collected. Some examples are provided below - you can customize the examples as required or create your own. There is more flexibility here than having a built in alert system, but it does require more effort to setup. Also, there is a risk that custom alerts could be broken by future schema changes to the repository database. DBA Dash might be extended at some point in the future to have built in alerting that can be configured using the GUI.