Caring about IT data quality is not a new concept. The discipline of Data Quality Management (DQM) has evolved from the transformation (or data cleansing) element of the traditional database Extraction, Transformation and Loading (ETL) process. To get a holistic view of an IT landscape requires most organizations to federate data from multiple sources and handle conflicts by deciding which source takes precedence. 10 years ago, the very notion of federating Configuration Items (CI’s) from multiple Configuration Management Databases (CMDBs) was ground breaking. However, CMDB projects continued to fail because it was necessary to intelligently merge multiple CI sources at an attribute level.
Today, we can handle federation of CI’s at an attribute level. Being successful in IT Service Management (ITSM) requires a clean CMDB containing purified data. Every major IT technology provider has their own favorite discovery solution and database. These pools of valuable asset data need to be consolidated to get a holistic view of your IT estate. Things get really interesting when you have an Managed Service Provider (MSP) managing a heterogeneous environment that you need to map for billing.
To understand the data problem a little better, consider for example, you are using an HP IT Asset Management (ITAM) solution to perform a discovery for a Windows server fleet. Can you rely on the CPU count it reports? Probably not. Microsoft’s System Center Configuration Manager (SCCM) is a more authoritative source for that attribute. Some organizations merge more than 10 sources and setup precedence rules for each. In fact, both 451 Research and EMA recently published reports that the number of sources IT deals with is increasing over time. The good news is that you only setup the rules once in order to automate the merge and reconciliation of your asset data.
At Blazent, our focus is on data quality, so we sponsored a recent 451 Research study (The State of Enterprise Data Quality: 2016) which compiled responses from 200 IT executives on their views of data quality.
One of the more enlightening sections of the report highlights at an attribute level. The four highest scoring desirable data attributes (with more than 50% congruence) were:
- Data integrity: Who can trust data with questionable integrity? Blazent discerns the most trustworthy source for every Configuration Item attribute to attain the highest data integrity.
- Accuracy: Some applications can suffice with a smidgeon of inaccuracy, but inaccuracy won’t help you ace an audit. Blazent’s data validation rules assure attribute level accuracy.
- Consistency: How do you see trends when every data point sets a new baseline. Blazent ensures your CMDB is continuously fed with the same source to deliver data consistency.
- Validity: You can extrapolate across a missing data element, but an invalid value will lead you astray. Bad data leads to bad results. Blazent checks for missing values and performs checks the data source and conflicts to verify validity.
Just to underscore my opening point, the research found that roughly one-third of respondents had some doubt about whether the data they were using was the correct data for their purposes.
You can read about this insight and many others by downloading the full report from http://goo.gl/V9wGNB