This guest blog was written by Nathan Camp of LogicMonitor.
Death by “Doh!Nuts!” is when a seemingly harmless, completely obvious, and avoidable issue goes unnoticed, swarms, and then crushes your software applications and the servers on which they reside. When you figure out why the database server crashed, you typically hear yourself shriek out, “Doh!,” and “Nuts!”
I absolutely love Liaison’s Delta and ECS software. It’s some of the most powerful on-premise data integration software out there. But, like all powerful integration software, the initial translation map building process requires human trial and error iterations to confirm that data output matches the expected format.
A best practice, available in many software applications and supported by Delta and ECS, is to initiate verbose logging during the development and testing phases of the design process, which can then be either turned off, or better yet, turned to only log errors and issue warnings.
And this is where the Doh!Nuts! come in.
Doh!Nuts! always cost Network and System Admins time in diagnosing, correcting, and restarting services, and often impact business partner relationships.
Some users would occasionally forget to turn off the verbose logging, especially when chasing the glory of going production-live with their systems. As you can already guess, they would go live, their software would diligently report all aspects about the successful message exchanges, and stuff these kernels of non-important data into the production database. At about one month in, the software would screech to a sudden halt and stop extremely important business-to-business data exchanges.
That’s where network performance monitoring solutions, like LogicMonitor, can watch both the hardware and applications running on them and warn in advance that a pending Doh!Nut! is about to hit. A brief list of a few key Windows SQL or MySQL database metrics that are monitored and alerted on include:
- Volume usage
- Total IO operations per day
- Disc throughput and latency
- Processor queue
- CPU usage
With loss-less database technology, the granular detailing allows predictions based on accurate trending metrics. You can see in advance when new disc volumes may need to be added to handle growing data, or better, when you may need to run your clean-up routines to remove non-important data. And with integrations to incident reporting and support ticket solutions like PagerDuty, ConnectWise, and several other leaders, LogicMonitor provides real-time, smart actions for IT rapid response.
Yet, that’s not all that can be done with advanced SaaS-based solutions. Mike Suding, one of LogicMonitor’s sales engineers, recently tested the boundaries of these performance monitoring and alerting systems. Mike took the approach that system alerts can be used as the basis to actually orchestrate a hands-free, “fix the problem” response. The below video shows a brief example of how he used network performance monitoring tools to clear out the temp files from the hard drive.
For more helpful ideas on monitoring and managing your IT infrastructure, you can subscribe to Mike Suding’s blog or his YouTube channel. Be sure to also check out LogicMonitor’s blog and video case studies list.