Monthly Archives: April 2012

Cloud-Capable “Real Time”

Posted on by
Reply

Our review of definitions for “real time” over the past couple of weeks has brought us to the point where we are now ready to answer the question: Is the cloud capable of real-time performance, and if so, how?  We have reviewed some of the technical definitions for “real time,” and considered some expert advice on a practical, realistic definition of “real time”.  Now let’s see whether or not the cloud really is capable of supporting real-time computing, according to these definitions.

To sum up E. Douglas Jensen at Time-Critical Technologies, real-time computing in the real world means achieving optimal system performance within given time constraints.  So, what is optimal system performance for a cloud system?  And what time constraints are we considering here?

Each cloud application has its own time constraints.  What seems uselessly slow for one set of users may be perfectly acceptable for others.  For example, in many business applications data that is just a few seconds old is considered real time.  In fact, many executives would boast about having access to up-to-the-minute data.  Managers and analysts running certain industrial applications like inventory control or end-of-shift reports have similar requirements.  When these users talk about real-time data in the cloud, delays of a few seconds or even minutes might be perfectly fine.

Fast moving lightsOn the other hand, most operators working at a control panel in a plant expect to see things happen immediately.  When they click a button, the light should come on right away, not after a few seconds, or even one second later.  Values should update as they change in the process.  Trend lines should be smooth curves drawn on the page, not jagged peaks that appear intermittently.  As far as we are concerned, a cloud system that claims to be real time should be able to emulate that experience very closely.  For our purposes, then, we can define “real time” for the cloud as follows:

“Real-time” cloud: Remote accessibility to data, with local-like immediacy.

Of course, the remote aspect of any cloud application will always have an impact.  The Internet and other networks inescapably introduce latencies into the data flow.  This kind of delay in delivering the data brings to mind the US Defense Department Military Dictionary definition of “real time,” which is: “Pertaining to the timeliness of data or information which has been delayed only by the time required for electronic communication. This implies that there are no noticeable delays.”

A real-time cloud system should have no noticeable delays, or at least, no more than absolutely necessary.  Any intermediate software running on the cloud should support high-speed data throughput.  Latencies should be no more than a few milliseconds over the network latency.  The infrastructure should almost certainly be data-centric, minimizing the need to convert between HTML, XML, SQL, or other data formats.

Broadly speaking, when people are working with the system, we can aim high and set a goal for our time constraints at human response time.  The user should feel like he or she is working on a local system.  Any extra processing time over and above network communication time should be kept to an absolute minimum.  And what about in M2M (machine to machine) applications?  Here again, there should be as little delay as possible beyond any networking latencies.

Although the debate about the definition of “real time” may continue for years to come, we can glean enough for our purposes.  For our real-world applications on the cloud, we can define “real time” as achieving optimal system performance within given time constraints.  So, if we define our time constraints to be human response time, and accept the limitations of networking latencies on system optimization, then we can confidently assert that the cloud is capable of supporting real-time systems.

Realistic “Real Time”

Posted on by
Reply

As we look over the technical definitions for real-time computing, it is clear that there is no way the cloud can support hard real time performance.  The cloud is not deterministic.  Despite advances in networking technology, there will always be delays and network breaks, making it impossible to implement any kind of hard real-time system where missing a deadline would cause a total system failure.  If we want to talk about real-time data in the cloud, we need to be clear on what we mean by real time.  We need a realistic definition of “real time” for the cloud.

Someone who has done a lot of thinking about how to define “real time” is E. Douglas Jensen.  Jensen has spent over 30 years developing real-time systems for industrial and military applications in institutions like Carnegie Mellon University, Hewlett Packard, and Honeywell.  More recently he has worked as a consultant for MITRE and the U.S. Department of Defense, as well as in his own consulting practice, Time-Critical Technologies (TCT).

According to Jensen, once you step away from the clear-cut definition of “real time” as hard real time, things get vague.  His experience is that researchers and academics working in the lab have created good theoretical models for real-time computing, but when we attempt to apply those theories in real-world situations, they are often not practical. This is particularly true in distributed systems, and the cloud is the ultimate distributed system.

On his website, Real-time for the Real World, Jensen offers a new approach to thinking about real-time systems in a realistic way.  First, he says that anything other than the concept of “hard real time” has been difficult to pin down and define, even by the experts. Second, it is tough to do any kind of experimental research on real-time systems in the real world because they are typically big, complicated, expensive, and mission-critical.  He then offers a reformulation of the idea of real time that has enabled him to construct highly complex real-time computing systems.

Jensen boils the issue down to two concepts: 1) time constraints (or deadlines) and 2) achieving acceptably optimal system performance within those constraints.  He says, “Real-time computing is about satisfying time constraints acceptably well with acceptable predictability—according to application- and situation-specific acceptability criteria, given the current circumstances.”

As in real life, time constraints can vary in immediacy and importance.  Some deadlines might be like catching a plane flight—not immediate, but important.  You may have some time to get to the gate, but when the plane takes off, you’d better be on board.  Other deadlines might be more like a phone call, which is always immediate but may not be important.  You get no advance warning when the phone rings, but if you don’t answer until the third or fourth ring, no problem.  Who is calling determines the importance, and sometimes even letting it go to voice mail may be just fine.

Also in real life, many things can happen at once.  While you are rushing to catch the flight, you realize that you left your ticket at home.  Just then your phone rings, someone bumps you from behind, you drop your bag, and then you notice that your wallet is missing.  What to do first?

Real-time computing systems are often put under similar demands, and need to respond in the best possible way.  An optimized system would meet all hard deadlines as often as possible, and minimize the number of soft deadlines missed and/or minimize the lateness of the response time.  Generally speaking, this is what it means to achieve optimal system performance within given time constraints.

A realistic approach to real-time establishes that each system, large or small, fast or slow, will have its own particular time constraints and acceptable level of optimal performance. Naturally, that can include cloud systems.  Taking this approach, we are now ready to look at the demands of some typical real-time systems, and how they can best be met when moved to the cloud.