Building and managing cloud applications is not easy. Teams come face to face with these challenges: agility, manageability, performance, scalability, continuous availability and of course, security. Join us for “The Agility Challenge: Powering Cloud Applications with Multi-Model & Mixed Workloads” webinar where we will deep dive into challenges customers face with multiple data models such as graph, mixed workloads and how DataStax Enterprise can help.
Video: https://youtu.be/1tKDxkexzFE
14. For Cloud Applications...
The Agility Challenge – Database Demands
Data Model Flexibility
Workload Isolation
Data Access Agility
Leverage Infrastructure Availability
Time to Market
Operational Ease
15. Solving The Agility Challenge
For Cloud Applications...The Solution: DSE
✔ Multi Model Support
✔ Multi Workload
✔ CQL, Spark, Search, Gremlin
✔ Multi Instance
✔ Developer Tools and Drivers
✔ Opscenter
Data Model Flexibility
Workload Isolation
Data Access Agility
Leverage Infrastructure Availability
Time to Market
Operational Ease
Good morning/afternoon and thanks for joining us.
Today we want to talk to you about your world of Cloud Applications, and how we can help you achieve the business success you hope to achieve.
We are going to focus today’s session on how DataStax Enterprise can help your business achieve by overcoming The Agility Challenge in the world of Cloud Applications.
We are going to focus on 4 main topics today.
1 – The Database Demands of Today
2 – How DSE Solves the Agility Challenge
3 – Highlight a couple of examples of how DSE customers have solved the Agility Challenge with DSE
4 – Review how DSE helps Cloud Applications drive business success
We live in an era that is changing the very fabric of who we are. The Digital Era is creating an opportunity for companies to impact a bigger and wider audience in more meaningful ways.
It wasn’t always this way though.
To understand and solve the digital challenge today’s successful companies concur, we will first take a look at the evolution of the database technologies.
The legacy database as we know it today got it’s start in the mainframe era of the 1970s. Data was very static and user access was extremely controlled. However, giving direct access to users to a computer was revolutionary back then, though the thought seems very simple to us today. These are the circumstances in the digital environment that gave birth to legacy databases.
The digital economy took a huge leap forward in the 1990’s with the advent of the Internet and along with it the introduction of client-server architectures. In the 1990’s most data was semi-static, mostly lists and reviews, with some user interaction. However, access to computers became more complex as multiple users could access a single server at the same time. In this era, relational databases grew in prevalence and became the defacto standard for data architectures.
Fast forward through the advent of mobile computing, social media, connected devices, the disruption on every industry at the hands of digital economy and you will start to appreciate the challenges of today’s modern architectures. We at DataStax call the modern data architectures Cloud Applications.
These Cloud Applications are expected to behave differently and perform at scale differently.
For e.g. if a company was launching an application 10 years ago from the company headquarters in California, it would think about the region and local users.
But today, companies cannot afford to think that way. Feeds, integrations and users for the app are going to be very distributed from the outset
So the end points for the app are going to be numerous and highly distributed.
TODO – Create example {If possible, give a personal example of an app you use that is highly distributed and making you interact with suppliers / people around the world and need immediate response and high performance, and yet get the right information to you based on data gathered and your interactions. Do not use Netflix as it comes later in the deck}
All this is possible today because of the cloud, in-house private data centers and very often - both. Leading to extremely complicated setups very quickly.
What we have just described are the Top 5 challenges of Cloud Applications:
Scalability
Performance
Availability
Security
Agility and Manageability
Today we are going to focus on Agility/Management
Agility in business requires a technical architecture that is quick and graceful.
Because of it’s birth during the mainframe era, Legacy Databases have become the opposite of Agile. Here’s an example showing the complexity, manual integration, and lots of points of failures.
As we walk through the rest of the presentation, we are going to highlight how DSE solves The Agility Challenge.
To level set, DSE is a purpose built database to meet the demands of Cloud Applications. DSE is a single, integrated platform that provides the features that are required for Agility.
At the foundation of DSE is Cassandra….
DSE offers a multi model platform capable of support KV, Tabluar, JSON, and Graph
DSE builds on top of the multi model platform through different lines of functionality, such as Search, Analytics, Unified Security, Multi-Instance...
Let’s start to decompose The Agility Challenge
We see and speak with customers constantly who are facing tough challenges in today’s Market.
Increased customer demand, through multiple channels
High demand for incredibly short cycles for financial results from technical investment
Large and/or rapidly growing user base’s require serious architectures spanning multiple physical and cloud based environments
These put pressure on the Owners, Designers, Developers, and Operators of Cloud Applications
How can one successfully meet these demands, this is The Agility Challenge
Our Customers tell us the Database is the central infrastructure point to help resolve TAC
They also tell us these database demands are key to solve the Agility Challenge
Data Model Flexibility -> Can our database handle multiple forms of data, natively?
Workload Isolation -> How can database interaction from one set of user demands, say end users, not impact another consumer of the same data set, say analysts.
Data Access Agility -> How can your database platform handle different data retrieval demands, straight lookups through simple queries, advanced searching, or deep analytics
Leverage Infrastructure Availability -> We’ve seen this countless times, HW procurement slows down the ability for teams to release new functionality. This issue compounds with specialized HW requirements from vertically scaled solutions.
Time to Market -> The promise of Agility is providing features to market as quickly as possible. Helping improve developer velocity is a big part of this item.
Operational Ease -> Being easy to operate in a geographically dispersed network is a key item for Agility with Cloud Applications
DSE is the Solution for the Agility challenge ….
We solve ...
When combined, the Database features for solving The Agility Challenge look like this, regardless of the type of application you are creating.
Let’s look at each feature individually now.
Large scale applications require consuming data in a number of different formats
And performance engineers want a quick key/value cache.
Analytics users want data in table format
Graph data is desired so that they can understand how users and machines interact
Web developers what data in JSON
The problem is that each user is attempting to solve their individual need with a separate database, and the problem with having multiple databases is how do you keep them in sync?
To do this you have to write a lot of complicated not very fun code.
How do I deal with failure scenarios, how do I deal with backup and recovery?
How do I do multi-datacenter replication?
How do I tune the database for performance?
This all requires lots and lots of people. These people need to become experts in each of these databases and that means additional time and cost.
We want to free your teams so that they can focus on functionality that matters to your users and company
How do we do that?
With multi-model support we let you interact with data the way that makes sense for your application. Write data once, and it's visible to everyone.
In a single platform we've combined the ability to interact with Key/Value, tabular, Graph and JSON data.
Even more important is that we've also got multi-workload capabilities so that your analytics users and application developers can use the same platform, the same data to satisfy transactional, search, and analytics workloads. Write data once, and it's available to all these teams in real-time.
Instead of becoming database experts we let you focus on what matters most to your users.
DSE’s multi-model and mutil-workload solutions solve The Data Model Flexibility and Workload Isolation demands of TAC
An Agile and Impactful User Experiences is directly related to how data is structured in a database.
Geographic specific requests, javascript optimized requests, highly connected or complex data relationships requests, analytical requests, etc are all examples of the types of data access that is required by today’s Cloud Applications. Providing a superb user experience means that a Cloud Application’s database must meet the Data Access item of TAC
DSE solves the Data Access demand of TAC by providing user access through:
CQL Queries for direct, extremely low latency, high concurrency lookups manipulation
Graph traversals provide an incredibly flexible way of navigating data relationships. We’ll show a bit more on this one in just a minute.
Fuzzy, type ahead, spell check, faceting, etc are all examples of how DSE Search, powered by Apache SOLR, provides flexible and fast access to data stored in DSE.
Aggregating data for deep analysis is key for today’s Cloud Applications. DSE provides this type of data access through our Analytics solution powered by Apache Spark.
Let’s take a deep dive look at how Graph Traversal helps solve TAC.
We hear from our customer's that their strict data hierarchies hamper their ability to provide an agile, response, and incredibly impactful user experience.
For example, we have a large customer who described a situation where their marketers identify new, competitive advantage for their UX, but their data is structured in the database like in a strict hierarchy. The marketers want to provide a new way to access product information and they need to release the change very quickly. Our customer stated that a large amount of effort is required to either [1] create a duplicate and complicated caching layer in the application to enable the application to control the structure of their data or [2] change the structure of the data in the database. Neither option is desirable.
Our customer is exploring the use of Graph traversals to solve this instance of the Data Access challenge.
DSE Graph storage + Graph traversals through the Gremlin query language provide an incredibly flexible way to access data within Cloud Applications. This means that a change in desired UX data navigation requires only a query change not a large, expensive, and time consuming restructuring of data in the data base or complicated front end application hacks to replicate the data in a caching layer.
This is an example of how DSE helps our users iterate on data access ideas to solve TAC.
TODO
Leveraging Infrastructure Availability
A lot of our customers have stated that their infrastructure choices limits their agility to respond quickly to scale events.
Infrastructure in a lot of enterprises are not agile and this poses a very real component of TAC
Being able to leverage existing investments in infrastructure or being agile with your architecture when you’re infrastructure procurement is not agile is what DSE 5.0 provides with Multi instance.
…
For some customers it's cheaper to use really large machines and run multiple instances of DSE.
With multi-instance support DSE intelligently provisions these multiple instances so that no two replicas of your data exist on the same physical hardware. This way if you lose a machine, it doesn't hurt your availability.
A key predictor of success if being First to Market with a new feature. Time to Market is directly impacted by Developer Velocity, meaning DV is a major demand of TAC.
DSE solves the Developer Velocity demand of TAC through a unified set of drivers across a wide range of native languages that provide secure access to all of the functionality of DSE
DSE provides graphical development tools such as Developer Center and DataStax Studio that help developers quickly build and deploy Cloud Applications.
DSE Drivers are offered in the natural languages used to build today’s Cloud Applications.
DSE Drivers act as “smart” clients within a Cloud Application and make cluster connectivity, load balancing, retry logic easy while at the same time ensure that client applications are leveraging the multi model and multi workload power and performance of the DSE platform.
DataStax Studio is a newly released developer tool. It has been introduced in DSE 5.0 and is an interactive tool for exploring and visualizing large datasets using DataStax Enterprise Graph.
DataStax Studio is being enhanced to provide the same great visual feedback of DSE Graph for the rest of the DSE suite.
Getting to Market quicker through increased development velocity is easy with DSE’s rich set of Development tools.
Operating a database cluster is critical for the Agility of a Cloud Application. The benefits of a multi model, multi workload environment can only be achieved through a well running cluster. DataStax’s Opscenter simplifies the management of operations for DSE by providing Monitoring, Provisioning, and Management Services.
From OpsCenter you can get a high level view of your DSE Cluster, see the overall health of your system or drill into individual nodes. Check to see whether all nodes are up and running green, under some heavy load, or if you have any failures. You can also configure proactive alerts to bring issues to your attention. For example, node failures, spikes in throughput, or latencies that don’t meet your SLAs.
Not only monitoring though, you can also administer changes to your cluster through OpsCenter, run best practice checks to make sure your configuration is optimal for performance, backup and restore your clusters, add more nodes to fuel your applications growth, and more.
OpsCenter allows you to manage your cloud application with ease, even while running at epic scale; allowing you to focus on your business value, not the administration of your architecture.
Opscenter is yet another way that DataStax solves The Agility Challenge.
The Agility Challenge is caused by
Increased customer demand, through multiple channels
High demand for incredibly short cycles for financial results from technical investment
Large and/or rapidly growing user base’s require serious architectures spanning multiple physical and cloud based environments
With DSE, the pressures on the Owners, Designers, Developers, and Operators of Cloud Applications are removed….providing an Agile data platform that powers business success.
Let’s now look at a few examples of DSE Customers and how they solved TAC.
One company that has solved The Agility Challenge with DSE is British Gas.
Background:
Connected Boilers are sending real-time data for preventative maintenance to identify abnormal spikes and readings, and proactively schedule a repair
They have 2+million smart meters with in-home displays (you can see your consumption while you are standing in the kitchen, and via mobile apps) - that are sending electricity readings every 10 seconds and gas readings every 30 minutes. And British Gas is using machine learning to disaggregate the energy data to identify which devices are consuming how much energy – since meters have an total consumption – for e.g. refrigerators have very cyclical patterns where they go on and off
Their Hive Connected Thermostats are streaming temperature time series data
British Gas uses multi model to store different in different formats, whether the key value style IoT data or the more tabular Analytical information used to compute outcomes.
British Gas uses DSE’s multi workload solution to ingest different streams of data within a highly available set of nodes, while computing deep analytics on another set of highly available set notes all within the same DSE cluster, all without having to preform any ETL to synchronize data across systems.
British Gas, a 200 year old company has an innovated Connected Homes business unit and are competing with NEST and giving Google a run for its money.
They are a great example of a DSE customer who has solved TAC with DSE.
Let’s look at another example of how a DataStax partner, Accenture, uses DSE to meet the multi model and multi workload demands of The Agility Challenge
Accenture has built a cyber security attack named Asgard.
Accenture's Security Graph Analytics for Real-time Defense
ASGARD is able to query billions of alerts from weeks – or years – of data and do it faster than ever before. It is a scalable, accelerated threat detection system that provides insights across enterprise data to help identify existing threats.
ASGARD is built upon DSE Graph for end user access and visualization. ASGARD leverages Spark Analytics to process and manipulate threat data as it’s entering the system. Thus giving Accenture the ability to solve The Agility Challenge for the clients of the ASGARD system.
Finally, here’s a quote from the Co-founder and CTO of PokitDok.
PokitDok is a cloud-based API platform designed to make healthcare transactions more efficient and streamline the business of health
PokitDok participated in the EAP program for DSE 5.0 to understand how DSE, with a focus on DSE Graph, can help PokitDok solve their Agility Challenge. The quote provided by Ted Tanner speaks for itself.
PokitDok has couple the power of DSE Graph with DSE Search to explore new ways of providing functionality for their users.
Full Quote:
“DataStax gives PokitDok the power to bring economic intelligence to the healthcare industry,” said Ted Tanner, Co-founder and CTO, PokitDok. “Graph theory and machine learning have vital roles to play in improving the system and it all comes down to the right tools to effectively harness data. DataStax allows our data scientists to be even more creative in novel solutions to healthcare Big Data by taking the developer operations engineering burden and by creating an effortless graph and indexing analysis stack. They have brought the ‘Easy Button’ to graph theoretic analysis.”
The global, complex demands of today’s Cloud Applications require Agility in the database layer to ensure business success.
DataStax Enterprise, with it’s multi-model, multi workload architecture solves the challenges and demands of today’s Cloud Applications.
DataStax Enterprise is the only solution to take this world – a chaotic, operationally complex and difficult patchwork of backend technologies and transform it – [next slide]
[Transform it into] a seamlessly, integrated, implementation that enables you to focus on your business, focus on your ability to deliver real-value through your cloud applications. DataStax Enterprise, with it’s secure, operationally simple, built from the ground up enterprise capabilities makes it the best fit for Cloud Applications and the right solution to solve The Agility Challenge!