Click here to close now.

Welcome!

Microservices Journal Authors: Elizabeth White, Liz McMillan, Blue Box Blog, AppDynamics Blog, Pat Romanski

Related Topics: Microservices Journal, XML, Oracle

Microservices Journal: Article

Improving the Efficiency of SOA-Based Applications

Using an Application Grid with large XML documents to build SOA applications that scale linearly and predictably

According to Moore's Law [1], processing speed and storage capacity have been doubling about every two years since the invention of the integrated circuit in 1958.

Yet it seems that our propensity for building larger more complex software systems that anticipate these improvements inevitably outpace the exponential growth in capacity to support these systems. SOA is becoming more broadly adopted, along with the practice of using XML as a means of communicating data between services and the more rapid adoption of applications to Internet scale. Staring you in the face of your application's success, the potential to overwhelm your systems has become very real, and may happen at times when you least expect it.

How do we get ahead of this trend? Given that memory and storage are always increasing in the realm of enterprise computing, software needs to keep up with the pace. We need to architect from the beginning using the proper approach toward achieving linear scalability with predictable latency. Data files and feeds are increasing in size, requiring more processing, and becoming more cumbersome to manage with software designed to materialize entire files before consuming them. In some cases, the operations that are to be performed require multiple input sources to be consumed before processing can begin.

Those who are building the eXtreme Transaction Processing (XTP) style of applications - such as Telco call setup and billing, online gaming, securities trading, risk management, and online travel booking - understand this challenge well. The broader use case that is applicable across more industries is web applications that need to scale up to Internet volumes, against backend systems that were never designed to handle that kind of traffic.

Boundary Costs
In discussions with customers about scaling a SOA with predictable latency, the term that often comes up is "Boundary Costs." To put this in context, consider the following scenario - an XML document that may have originated from an internal application, database, an external business partner, or perhaps converted from an EDI document, needs to be processed by a number of services, which are coordinated by a BPEL process or an ESB process pipeline. The common approach is to place the XML document on the bus and have the bus invoke the services in accordance with the process definition, passing the XML document as part of the service request payload. Each service that needs to process that data will access the XML accordingly. Interaction with a database may also occur. This approach, as illustrated in Figure 1, sounds simple enough.

Figure 1: Calling services using BPEL process or Service Bus pipeline

However, in practice there are challenges to scalability when using this approach. What is the cost of crossing the boundary from one service to the next? How many times does that cost get incurred in the context of invoking a simple business process? What if the XML document is really large in the multi-megabyte range, or there are lots of them numbering in the thousands, or both?

Compounding this challenge is the reality that most IT environments are a mixture of platforms and technologies. Regardless of how efficient your process engine or service bus might be, the processing at the service endpoint might still become a bottleneck. A recent conversation at a customer site revealed a 15-step business process that normally takes 15 seconds to run, but of late under peak loads it is violating its 30-second SLA. The developers had spent the better part of the past two years optimizing and tuning every last bit of performance out of each one of those 15 services, and the remaining culprit identified for the poor end-to-end latency is the boundary cost between the services. A detailed examination revealed that each of the 15 service calls was spending 1-2 seconds in an open source web service toolkit doing parsing and marshaling of the XML payload. This is not intended to be a disparaging comment about open source web services toolkits, but is simply illustrating the point that parsing and marshaling of XML at the endpoints can introduce latency that can add up pretty quickly.

As illustrated in Figure 2, each service invoked needs to read the XML payload from its on-the-wire serialization form, and parse the XML into a native Java or .NET object form to be processed by the business logic. In addition if database interaction is required, then there is an additional object to relational mapping that needs to occur. Finally, the inverse of those steps needs to occur in order to generate a response to the service request and send that along to the next downstream service in accordance with the business process that is coordinating the interaction between the services.

Figure 2: Service request boundary cost between XML to Object to Relational and back again for each invocation

A popular approach for dealing with XML in a SOA is to use web services and XMLBeans. Using XMLBeans, objects are typically created by fully materializing the inputs and outputs, as this allows for maximum usability and processing. In-memory processing may include sorting, filtering, or aggregation operations, all of which increase the overall memory required to deal with each call. This strategy is not scalable and cannot be applied to many of the use cases in this area. Many products support streaming of XML, but this may limit the ability to do anything meaningful without putting the data somewhere else first.

What if there was a way to take this information and store it in an application grid, a place where the size of the data and the processing capability can far eclipse that of any single machine or process? The application grid can utilize the combined memory and processing power of multiple machines in order to complete an operation, such as the application of a complex formula or filter across an enormous data set. The application grid also provides the ability to hold the data for longer periods of time beyond the cycle of a single service request, survive server restarts, and even work across network boundaries.

If we could combine the power of the grid for data storage and manipulation with the efficiency of streaming, the result would be a highly scalable system capable of processing much more information than before. Using a combination of complementary technologies here, we achieve our goal of spreading compute operations across a distributed network of machines, and we lessen the processing and memory requirements of our data consumers - SOA services, application servers, and client applications. We also remove the need to use a database for intermediate storage of data while it is (or simply so it can be) processed. By using an application grid we can also implement patterns where we pass around references to data, rather than the data, resulting in huge efficiency gains in the communications layer, and dramatically reducing or eliminating the boundary cost.

This article includes a code example that covers the use case of processing large XML files in an application grid. In a typical XML file, there are a usually elements that repeat without any pre-determined limit. Using a STAX parser to handle streaming XML, and JAXB to handle conversion between XML and Java objects, we can extract these repeating elements from the XML stream and put them on the application grid as individual objects. The implementation can populate the grid with these objects, and do so with a limited amount of memory consumption. Once populated, the grid can process the data across the multiple machines that constitute the grid. Each grid member processes an operation or a filter and passes intermediate results to the grid client, which then assembles them into a final result set.

What Is an Application Grid?
An application grid is a horizontally scalable agent based in an in-memory storage engine for application state data. This effectively provides a distributed shared memory pool that can be linearly scaled across a heterogeneous grid of machines that consists of any combination of high-end and lower-cost commodity hardware. Use of an application grid in an application simultaneously provides performance, scalability and reliability to in-memory data.

One way that an application utilizes an application grid is to use API-level interfaces that mimic the Java Hashmap, .NET Dictionary, or JPA interfaces. An alternate approach is to use a service-level interface from a SOA environment. As applications or services place data into the application grid, a group of constantly cooperating caching servers coordinate updates to data objects, as well as their backups, using cluster-wide concurrency control.

As shown in Figure 3, the request to put data to the map is taken over by the application grid and transported across a highly efficient networking protocol to the grid node P, which owns the primary instance data. The primary node in turn copies the updated value to the secondary node B for backup, and then returns control to the service.

Figure 3: Application grid clustering ensures primary / backup of in-memory data on separate machines.

The application grid stores data across multiple machines with complete location transparency as it sees fit. A unique hash key value is all that is necessary to retrieve the stored data at a future point, regardless of where the application grid chose to store the data. This prevents the application logic from dealing with complex location dependencies and manual partitioning schemes. If one or more nodes in the grid fails, or can't be reached due to network failure, the application grid will immediately react to the failure and rebalance the data across the remaining healthy nodes. This can happen even if the failing node had been participating in an autonomous update operation. In Figure 4, the primary owner ‘P' of a piece of data fails while in the midst of retrieving data for the service. The get() request is immediately routed to the backup node and a new primary / backup pair is allotted.

Figure 4:  Application grid provides continual failover of in-memory state data

This data stored in the grid can be anything from simple variables to complex objects or even large XML documents. In our case we chose to fragment what would have been very large XML documents into smaller parts and store those XML fragments as Java objects in the application grid. This allows us to do parallel queries against the data using the Java APIs.

The application grid supports a range of operations including parallel processing of queries, events, and transactions. For large datasets, an entire collection of data may be put to the grid as a single operation, and the grid can disperse the contents of the collection across multiple primary and backup nodes in order to scale. In more advanced applications, the grid may even execute business logic directly and in parallel on data storage nodes, and do so with data and logic affinity such that the logic executes on the same machine that is storing the data that the logic is operating on.

More Stories By Dave Chappell

David Chappell is vice president and chief technologist for SOA at Oracle Corporation, and is driving the vision for Oracle’s SOA on App Grid initiative.

More Stories By Andrew Gregory

Andrew Gregory is currently a Sales Consultant at Oracle Corporation. He has worked in Development, Product Support, Infrastructure, and Sales over 13 years in the industry.

Comments (1) View Comments

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.


Most Recent Comments
jhv1blz5 07/03/09 10:31:00 AM EDT

The article validated SOA as an IT architecture paradigm that can be leveraged in many ways. Taking data storage, scalability and application performance to a nifty level using SOA Application Grid infrastructure will no doubt enhance data and application performance on Oracle architecture platforms, it also has the promise of a cost effective and efficient IT delivery model. The very benefits of SOA.

@MicroservicesExpo Stories
The 17th International Cloud Expo has announced that its Call for Papers is open. 17th International Cloud Expo, to be held November 3-5, 2015, at the Santa Clara Convention Center in Santa Clara, CA, brings together Cloud Computing, APM, APIs, Microservices, Security, Big Data, Internet of Things, DevOps and WebRTC to one location. With cloud computing driving a higher percentage of enterprise IT budgets every year, it becomes increasingly important to plant your flag in this fast-expanding bu...
What are the benefits of using an enterprise-grade orchestration platform? In their session at 15th Cloud Expo, Nate Gordon, Director of Technology at Appcore, and Kedar Poduri, Senior Director of Product Management at Citrix Systems, took a closer look at the architectural design factors needed to support diverse workloads and how to run these workloads efficiently as a service provider. They also discussed how to deploy private cloud environments in 15 minutes or less.
Cloud Expo New York is happening from June 9 - 11. This event brings together the worlds of Cloud Computing, DevOps, IoT, WebRTC, Big Data and SDDC. We hope to see you there-members of the Blue Box team will exhibit in booth 218 next to the DevOps area. Plus, our Chief Product Officer, Hernan Alvarez, will present his talk "The Cloud Has a Down-and-Dirty Lining" as part of the Operations track in the DevOps Summit portion of the event on June 9 at 11 am. Learn more about his session her...
Docker is an open platform for developers and sysadmins of distributed applications that enables them to build, ship, and run any app anywhere. Docker allows applications to run on any platform irrespective of what tools were used to build it making it easy to distribute, test, and run software. I found this 5 Minute Docker video, which is very helpful when you want to get a quick and digestible overview. If you want to learn more, you can go to Docker’s web page and start with this Docker intro...
The 5th International DevOps Summit, co-located with 17th International Cloud Expo – being held November 3-5, 2015, at the Santa Clara Convention Center in Santa Clara, CA – announces that its Call for Papers is open. Born out of proven success in agile development, cloud computing, and process automation, DevOps is a macro trend you cannot afford to miss. From showcase success stories from early adopters and web-scale businesses, DevOps is expanding to organizations of all sizes, including the...
Over the years, a variety of methodologies have emerged in order to overcome the challenges related to project constraints. The successful use of each methodology seems highly context-dependent. However, communication seems to be the common denominator of the many challenges that project management methodologies intend to resolve. In this respect, Information and Communication Technologies (ICTs) can be viewed as powerful tools for managing projects. Few research papers have focused on the way...
As the world moves from DevOps to NoOps, application deployment to the cloud ought to become a lot simpler. However, applications have been architected with a much tighter coupling than it needs to be which makes deployment in different environments and migration between them harder. The microservices architecture, which is the basis of many new age distributed systems such as OpenStack, Netflix and so on is at the heart of CloudFoundry – a complete developer-oriented Platform as a Service (PaaS...
17th Cloud Expo, taking place Nov 3-5, 2015, at the Santa Clara Convention Center in Santa Clara, CA, will feature technical sessions from a rock star conference faculty and the leading industry players in the world. Cloud computing is now being embraced by a majority of enterprises of all sizes. Yesterday's debate about public vs. private has transformed into the reality of hybrid cloud: a recent survey shows that 74% of enterprises have a hybrid cloud strategy. Meanwhile, 94% of enterprises a...
There is no question that the cloud is where businesses want to host data. Until recently hypervisor virtualization was the most widely used method in cloud computing. Recently virtual containers have been gaining in popularity, and for good reason. In the debate between virtual machines and containers, the latter have been seen as the new kid on the block – and like other emerging technology have had some initial shortcomings. However, the container space has evolved drastically since coming on...
Enterprises are fast realizing the importance of integrating SaaS/Cloud applications, API and on-premises data and processes, to unleash hidden value. This webinar explores how managers can use a Microservice-centric approach to aggressively tackle the unexpected new integration challenges posed by proliferation of cloud, mobile, social and big data projects. Industry analyst and SOA expert Jason Bloomberg will strip away the hype from microservices, and clearly identify their advantages and d...
In her General Session at 15th Cloud Expo, Anne Plese, Senior Consultant, Cloud Product Marketing, at Verizon Enterprise, focused on finding the right mix of renting vs. buying Oracle capacity to scale to meet business demands, and offer validated Oracle database TCO models for Oracle development and testing environments. Anne Plese is a marketing and technology enthusiast/realist with over 19+ years in high tech. At Verizon Enterprise, she focuses on driving growth for the Verizon Cloud platfo...
With major technology companies and startups seriously embracing IoT strategies, now is the perfect time to attend @ThingsExpo in Silicon Valley. Learn what is going on, contribute to the discussions, and ensure that your enterprise is as "IoT-Ready" as it can be! Internet of @ThingsExpo, taking place Nov 3-5, 2015, at the Santa Clara Convention Center in Santa Clara, CA, is co-located with 17th Cloud Expo and will feature technical sessions from a rock star conference faculty and the leading in...
The widespread success of cloud computing is driving the DevOps revolution in enterprise IT. Now as never before, development teams must communicate and collaborate in a dynamic, 24/7/365 environment. There is no time to wait for long development cycles that produce software that is obsolete at launch. DevOps may be disruptive, but it is essential. The DevOps Summit at Cloud Expo – to be held June 3-5, 2015, at the Javits Center in New York City – will expand the DevOps community, enable a wide...
How does one bridge the gap between traditional enterprise storage infrastructures and the private, hybrid, and public cloud? In his session at 15th Cloud Expo, Dan Pollack, Chief Architect of Storage Operations at AOL Inc., examed the workload differences and required changes to reuse existing knowledge and components when building and using a cloud infrastructure. He also looked into the operational considerations, tool requirements, and behavioral changes required for private cloud storage s...
Cloud Expo, Inc. has announced today that Andi Mann returns to DevOps Summit 2015 as Conference Chair. The 4th International DevOps Summit will take place on June 9-11, 2015, at the Javits Center in New York City. "DevOps is set to be one of the most profound disruptions to hit IT in decades," said Andi Mann. "It is a natural extension of cloud computing, and I have seen both firsthand and in independent research the fantastic results DevOps delivers. So I am excited to help the great team at ...
Software is eating the world. Companies that were not previously in the technology space now find themselves competing with Google and Amazon on speed of innovation. As the innovation cycle accelerates, companies must embrace rapid and constant change to both applications and their infrastructure, and find a way to deliver speed and agility of development without sacrificing reliability or efficiency of operations. In her Day 2 Keynote DevOps Summit, Victoria Livschitz, CEO of Qubell, discussed...
How can you compare one technology or tool to its competitors? Usually, there is no objective comparison available. So how do you know which is better? Eclipse or IntelliJ IDEA? Java EE or Spring? C# or Java? All you can usually find is a holy war and biased comparisons on vendor sites. But luckily, sometimes, you can find a fair comparison. How does this come to be? By having it co-authored by the stakeholders. The binary repository comparison matrix is one of those rare resources. It is edite...
There’s a lot of discussion around managing outages in production via the likes of DevOps principles and the corresponding software development lifecycles that does enable higher quality output from development, however, one cannot lay all blame for “bugs” and failures at the feet of those responsible for coding and development. As developers incorporate features and benefits of these paradigm shift, there is a learning curve and a point of not-knowing-what-is-not-known. Sometimes, the only way ...
Working with Big Data is challenging, especially when decision makers depend on market insights and intelligence from your data but don't have quick access to it or find it unusable. In their session at 6th Big Data Expo, Ian Khan, Global Strategic Positioning & Brand Manager at Solgenia; Zel Bianco, President, CEO and Co-Founder of Interactive Edge of Solgenia; and Ermanno Bonifazi, CEO & Founder at Solgenia, discussed how a revolutionary cloud-based BI along with mobile analytics is already c...
Hardware will never be more valuable than on the day it hits your loading dock. Each day new servers are not deployed to production the business is losing money. While Moore's Law is typically cited to explain the exponential density growth of chips, a critical consequence of this is rapid depreciation of servers. The hardware for clustered systems (e.g., Hadoop, OpenStack) tends to be significant capital expenses. In his session at Big Data Expo, Mason Katz, CTO and co-founder of StackIQ, disc...