I have selected a polling time of early evening.  In hindsight, I should have chosen 11::30 PM.  Maybe someday the time will get changed to this later time. Initially, decide on the criticality of the repairs. I am making the critical repairs before backups.  Less urgent repairs get pushed to the morning. Where I work the network, engineers have their hands in many tasks that are not their wheelhouse, like DBA tasks.  I have this running in the evening to keep an eye on them and their tricks.  Removing rights is not an option. Data is sent in Sunday through Thursday, only five days a week.

What am I looking for in the results?

  • Unexpected changes that could impact the users like large log files or waits types that need attention.  Can these be fixed in the morning while spending 10 minutes n each server?
  • Early warning of SQL agent jobs not completed for one reason or another.  For example, SQL agent jobs not ran.
  • Changes made by another admin that does now know what they are doing.  This type of change is a mixed bag, and when possible, they can wait until the morning.  The more critical, I take care of right away.
  • From time to time, I will get index recommendations.
  • Once a week, constant care gives me a more detailed report.
  • On a typical week, I get 28 emails a day and disregard 26 of them with no action required.  It is good that the systems are in this good shape.  The other two or so a day get emailed to work and processed in the morning if the issue can wait that long for action.

The short story for the use is the redundancy of what I normally do during the day.

How can this be used? How about reporting to your boss?  If they are like my bosses, they don’t fully understand the role of the DBA and what is truly at risk if something goes south with a database.  I thoroughly suggest you work on building a reporting tool out of this data and others.

You combined the Constant Care data in the ticketing, service request systems, and any other data sources you have to do your job.  One solution is to create a ticket in your ticketing system, but that takes a lot of time away from your job. Do you have time for that extra step?  Adding to the ticketing system would change my morning #DB Health task from 30-60 minutes to 180 minutes.  And would that show proper value to the tools you are using? Probably not.

Reporting in Power BI is an option, but it takes time and skill-building.  Power BI is a good option for later.  Storing the data in a database is an option. Again it takes time and needs to have some reporting tools for it.  Excel is a good option, and you can scale it to work with Power BI.

Initially, I wanted to keep it simple with a report like this for the monthly:

Month:
Server Alerts Ignorable Actions taken Notes
<Name> # # # #

Or

Server Tickets Service Requests Changes CC Alerts CC Alerts Ignored CC Action CTK Health CTK
<Name> # # # # # # # #

 

The yearly report  would be:

Jan Feb …. Dec
Server Alerts Actions Ignored Alerts Actions Alerts notices Actions Ignored
<Name> # # # # # # # # #
Notes

These are not perfect but a beginning.  It only accounts for constant care, tickets, and service requests.  Is this granular enough for your boss?

But this is what you have to figure out with your boss.  We are here to help you create this report getting your team to report to this level.

 

Privacy Preference Center