How to Monitor Oracle or MS SQL Databases

Here at DSP where we manage a lot of Oracle and Microsoft Databases on behalf of our customers. Quite often, when we on-board a new client, we find they have their own Monitoring Platform they use for overall Infrastructure Management.

This could include:

- Network Monitoring – port utilisation, bottlenecks and throughput
- Storage and Server – IOPs, CPU performance, capacity and availability
- Application and Database Monitoring – Wait times, utilisation, availability

Commercially available packages include IBM’s Tivoli, HP’s OpenView (or HP BTO), SolarWinds, Nagios, to hundreds of other less known brands that do just as good a job.

The challenge is not necessarily choice, or even technology. It’s the level of effort an IT operation has to spend integrating this tool into their infrastructures. Not only time and effort, but more importantly cost! Packages – to get the tool working how you want it to, can cost millions.

What tends to happen is an organisation would choose and deploy a package, then just use the basics of what it can do.

At DSP, we concentrate on the Database Layer and how it connects to the Application and the Infrastructure. Being specialists in MS SQL and Oracle we have painstakingly customised our Monitoring tool to focus purely on what a DBA needs to see.

Whilst we monitor the Infrastructure, we do so in terms of how the Database impacts it.

And whilst we monitor interfaces and ETL processes, we do so because of the database.

And when we monitor the Operating System, we do so to ensure we can support the database effectively.

Therefore, to summarise, our Monitoring tool is honed – over the course of a decade - to ensure customers do not have to go through the pain of choosing, using, maintaining and customising a clunky Infrastructure plug in. That they have a monitoring platform designed and customised by DBA’s, for DBA’s. All the checks, alerts, thresholds, false/positives etc. have already been done by us for most versions of Oracle or MS SQL – so our customers don’t have to.