Alerting on Availability, Space, Deadlocks, Backups, CPU – Essential Alerting for DBAs

by May 11, 2018

Precise tests availability, space, deadlocks, backups, jobs, CPU, load and performance trending plus many more conditions on a scheduled basis with defined blackout times.  When there is a problem, Precise sends an email, a message, executes a script, and/or sends a MIB to a ticketing framework.  Trend alerts are your early warning system.  One screen shows the condition of the database instances.

Precise helps DBAs minimize MTTR (Mean Time To Resolution) and better yet avoid alerts altogether.  Appropriately scheduled maintenance can help avoid contention and locking.  The best alert is the one that is avoided.  Our philosophy is that alerts should be meaningful and actionable.  We believe that the worst alert is one that DBAs ignore.

Precise is delivered with all, but availability, de-activated.  Only the alerts that are appropriate for an environment should be activated.  The alert set can be propagated to similar environments.

SQL Server availability alerting with history:

"

Critical alerts are thrown for group of database instances:

"

Custom alerts can be incorporated into Precise’s alerting process.  Alerts work with exception reports.  The idea is to show the trend over time.  Finally, automatic corrective action can be fired on a threshold breach.  You might test additional conditions to diagnose further or attempt to solve a problem, perhaps restart a service.