Saturday, April 25, 2009

Configuration Management Data Base (CMDB) and BTM

Any organization that has or is planning to implement the ITIL methodology will find great value in a Business Transaction Management (BTM) solution. BTM contributes significantly to the population and utilization of your CMDB, along with the auto creation of a service model relationship within your CMDB.

CMDB Population: Business Transaction Management solutions auto discover your application dependency map by monitoring the real transactions that run through an application's full topology. Additionally, the auto discovered transaction types are then added to the CMDB. Anyone accessing the CMDB that wants to see more specific information about various transactions will be referred directly to the BTM tool's transaction repository.

Enhanced Utilization: BTM solutions enable better utilization of a CMDB by linking true business activities to IT processes, enabling the management of IT from the business perspective. Additionally, BTM solutions connect specific transaction segments and servers – virtual or real – to service degradations enabling rapid resolution.

BTM Enables the Real Time Population of Your CMDB

BTM solutions are designed to operate around the clock in production environments and collect important data on all transactions that flow through all of the IT components in the datacenter. This enables your CMDB to be continuously populated with the critical business processes that are linked to their underlying IT counterparts.

Some examples of Configuration Items (CIs) that a BTM solution can populate a CMDB with are:

Business services (transaction types)

Data flows - the full structure of the infrastructure that supports each transaction - down to the methods that are executed - enabling much more granular impact analyses in change management

The Topology of each transaction type (true service models)

IP addresses of clients and servers

All of the possible physical servers that transactions may run through

All of the possible virtual servers that may be brought up on the fly in response to increasing load

The advantage of defining business relationship CIs with the help of a BTM solution is that those relationships are 100% accurate and are based on the flow of real transaction activations in the production environment, as opposed to other CMDB populating tools which must rely on assumptions, or manual input.

It is now possible to understand which changes affect which transactions – or business services - so that the criticalities of each change can be better understood from a true business perspective.

CMDB and the Help Desk Ticket

A BTM solution will automatically detect slow or hanging business transactions and open an Incident. By doing so, the IT Service Support or Help Desk team can be proactive in dealing with the incident. They can proactively start working on the root cause of the problem and fix it – possibly even before any users feel the negative impact of the degradation.

Tickets that are opened by the help desk are automatically put in the context of the CMDB since the BTM solution links the IT components that are involved in the transaction degradation to that ticket. This not only enables tickets to be sent to the appropriate administrator for resolution, it also enables that administrator to utilize CMDB information to help resolve the issue.

BTM – CMDB Scenarios

A Service degradation

The BTM solution identifies the degradation of a specific transaction type’s SLA

An incident is created and a help desk ticket is opened - flagging the server which is causing the latency - for example – the latency is due to a specific application server

The Help Desk forwards the ticket to the Application Team as part of the problem resolution process

The Application Team can now look at the change history of that specific application server within the CMDB and identify the cause of the problem

A Change is Being Contemplated

A change to a specific application in the application server is being contemplated

The Change Manager identifies all of the real business transactions that are associated with the change

The Change Manager will be able to anticipate the business impact that the change will have on the business users and will later be able to measure the impact with real user measurements

Prioritization of Incidents

Two servers fail at the exact same time

The system administrator knows which server is more critical to the business and can prioritize the recovery - something that is not possible to perform without the true model

Change Verification

Measuring the impact of a change that has been conducted

A change to the application has been implemented

The BTM solution will show the impact of that change by comparing the performance of transactions and transaction segments before and after the change

Roll Back Has Been Performed

The performance of all transactions in the current time frame is compared to the time frame of the last known consistent state in order to verify the roll back

Comparisons

The performance of specific transaction segments on two identical servers is compared in order to verify the consistency of performance

The topologies of two similar applications are compared in order to validate the quality of implementation

A CMDB for the Cloud

Cloud Computing's value proposition to the enterprise is a substantial one - maintaining a CMDB for the cloud introduces new challenges. These challenges are related to the real time scalability of applications in the cloud.

Business Transaction management solutions can at any given moment provide a real time snapshot of the cloud enabling a CMDB to keep up with the constant change.

Additionally, BTM's ability to effectively support a federated environment contributes to the stability and scalability of a CMDB in the cloud.

The CMDB can define a template that is fed in by the Business Transaction Management solution. In this manner the current status of the cloud can be shown – this can only be done with a real time BTM tool.

BTM Brings Value to Your CMDB Investment

Consider the following scenario; an application utilizes twenty different servers within the datacenter, a specific business critical service that the application provides really only utilizes only four of those servers. When building a CMDB with traditional tools it is nearly impossible to crack this scenario since most models are based on theories and assumptions. The only way of achieving this accurate level of granularity with the CMDB – short of going line by line through the application’s code - is by using BTM to help populate the CMDB with real data.

Implementing a CMDB is not a small investment, not to mention maintaining it, utilizing it, keeping it up to date and ensuring its validity. Business Transaction Management solutions enable organizations to better utilize and maintain their CMDB by filling in the gaps that occur in real time. BTM solutions are the only solutions that can link the degradation of a business service to the problematic IT component. This enables the location of the relevant CMDB data needed in order to resolve the problem.

Friday, April 10, 2009

BTM and Real User Measurements (RUM)

A complete BTM solution cannot ignore the “first mile” – the segment from the end-user to the data center. Real User Measurements (RUM) should be part of any complete BTM solution, since the origin of every user related transaction, is at the user.

Enterprises must have the ability to measure the level of service that their users are actually receiving - from their own desktop - and provide fast answers when performance degradation occurs.

Whether response times are degrading, or transaction failures are proliferating, IT staff must provide fast answers.

RUM must address two separate needs:

  • Web based applications used by remote users at their home or on a mobile device
  • Internal corporate users with desktop or Citrix based applications

For home users, installation on the home user desktop is usually not an option. BTM must provide the technology for measuring latencies from the home user’s browser, without installation on the desktop.

Enterprise users, tend to use numerous applications, some based on fat clients (no browser) installed on their desktop. In such cases a local agent has to be deployed to capture every transaction issued by the user in order to measure its latency and successful completion.

Agent based RUM solutions also provide inventory metrics – how many applications are actually being used, what applications are installed and executed within the user’s operating system, and so forth.