 |
 |
|
Production Data Masking |
Sourceless Performance Improvement |
Business-aligned Server Prioritization |
|||||
Ever since
the early days of IT, (aka Electronic Data Processing), it has been
commonly accepted to allow a certain percentage of IT Staff to have
access to the production environment.  These "trusted employees" were carefully screened and were usually in close proximity to management due to the confidentiality of critical sensitive data. Originally, this was a practical matter and was voluntarily implemented by the enterprise. Over the years, the onslaught of international Data Privacy Legislation has made this a compliance matter as well. Today's large, multi-national enterprise is faced with numerous cross-border data privacy exposures. Additionally, the deployment of third-party contractors, there is further separation from the traditional "trusted employee". It is now critical to evaluate solutions that will perform data masking in Production. This will allow the critical nature of Production Support to perform the necessary work, but not at the risk of exposing critical sensitive information in the process.
|
Quite
often, when application performance is suffering, the answer is
"Rewrite the SQL". But this is not as easy as it
sounds...especially when you don't have access to the source code. Customers of 3rd-party ERP/CRM applications such as PeopleSoft, SAP, Oracle e-Business, Siebel, Amdocs, Clarify and others, are frequently faced with the dilemma of knowing what is wrong, and knowing what the solution is, but find themselves at the mercy of the vendor to provide the fix on a timely basis, and something that won't break the budget. Other culprits include BI/DWH environments including Business Objects, Cognos, Crystal Reports and other end-user query and reporting tools. Sometimes, no matter how you specify the request in the tool, the SQL that gets generated is sub-optimal. And what about those Development Tools like Toad, DBArtisan, PL/SQL, SQL*Plus and others? Quite often the developers and DBAs that use these tools do not understand the implications of running tasks that cause massive performance problems, both in Production and in Test.
|
We have
all been there...on the phone with the help desk, when the agent says, "Sorry,
but my system is slow today...there is a lot going on in the system right now....happens
every day at this time" Why? A peak processing period is occurring and it is impacting the SLA on the critical business applications. The peak processing from processes outside of the critical business application, whether its system maintenance, a flood of BI requests or a large batch submitted during prime time. What is missing is a prime but basic function that can manage the resource allocation of the active sessions and running processes. Typically server resources are allocated on a first-come, first-served basis and will attempt to share resources equally across all applications. Unfortunately, this is not the way to go. Session and process resources should be allocated based on stack-ranking of business priority. This will align resource allocation with the business.
|