Let’s talk a little bit about data silos. A real-world silo, of course, is a farm tower used to store grain for future use or sale. They are usually towering buildings containing only one type of raw material. The concept of a silo serves as a metaphor for large collections of raw data that are typically stored separately from other raw data.
Servers and devices often silo data. Various machines store data, but not all of it is shared with other devices. Applications generate and store data, but only some applications…Might be so…if a well-written API (Application Programming Interface) is used. Over time, organizations find themselves with vast amounts of data, most of which is isolated, metaphorically stored in separate silos, and becoming part of a larger whole. there is no.
How edge computing can bring the perfect storm to your data silos
Data silos are a natural occurrence when it comes to enterprise networking, especially when it comes to edge-to-cloud networking. Every device at the edge produces data, but much of that data can remain on that device, or at least the cluster of devices at that edge location. The same is true for cloud operations. Data is created, stored, and sometimes exchanged in various cloud providers, most of which are isolated from the rest of the enterprise.
Also: How edge-to-cloud migration is driving the next phase of digital transformation
But when the right people and systems have access to all data across the enterprise, insights and actionable strategies emerge. Let’s look at one example that might occur with Home-by-Home, the fictitious home goods retailer mentioned earlier.
Home-by-Home sells a wall light fixture that mounts to the wall using a plastic bracket. Usually it’s a great seller. But every March and April, the company gets a ton of returns because the brackets crack. Returns come from all over the country, from Miami to Seattle. This is our first data set and is known to the stores themselves.
Brackets are assembled at the partner company’s factory. Factories typically run in temperatures above 62 degrees Fahrenheit, but in January and February the factory’s ambient temperature averages as low as 57 degrees. Here is his second data cluster, the temperature in the factory.
Neither dataset is connected to each other. But I looked into it a while ago and found that some plastic manufacturing processes start failing below 59 degrees. Without being able to correlate the factory data set with store return statistics, the company wouldn’t know that the slightly cooler factory was producing substandard brackets.
But by taking all the data and making the data set available for analysis (and AI-based correlation and big data processing), insights become possible. In this case, Home-by-Home has made digital transformation part of its DNA, so the company has been able to link factory temperatures to returns, and now customers who buy these luminaires experience far fewer failures. is becoming less
Data is everywhere, but is it actionable?
This is just one example of the possibilities of edge-to-cloud data collection. There are some important interrelated ideas here.
Your data is everywhere: Nearly all computers, servers, Internet of Things devices, phones, factory systems, branch systems, cash registers, vehicles, software-as-a-service apps, and network management systems are constantly generating data. As new data is generated, some of it is erased. Some of it accumulates until it clogs the storage device due to overuse. Part of it resides in a cloud service for each login account you have.
Your data is segregated: Most of these systems do not communicate with each other. In fact, data management often takes the form of considering what data can be removed to allow more data to be collected. Some systems have APIs for exchanging data, but most are not used (and some are overused). When my father complained about local business, he liked to use the phrase, “The left hand doesn’t know what the right hand is doing.” Organizations are just like that when data is segregated.
Correlating multiple inputs gives insight. A single data set can be subjected to comprehensive analysis to gain insight, but if data from one source can be correlated with data from other sources, trends are much more likely to be identified. increase. We have previously shown that factory floor temperatures have a distant but measurable relationship with the volume of returns in stores nationwide.
To do that, all data must be accessible across the enterprise. But these correlations and observations are only possible when analysts (both human and AI) have access to many data sources and know what it tells them.
Make data usable and turn it into intelligence
The challenge then is to make all that data available, collected, and then processed into actionable intelligence. To do that, we need to consider four things.
Initially tripData must have a mechanism to move from all these edge devices, cloud services, servers, etc. to where it can be acted upon. aggregatedTerms like “data lake” and “data warehouse” refer to this concept of data aggregation even though the actual storage of the data is fairly distributed.
Also: Edge-to-cloud digital transformation comes to life in this scenario for a major retailer.
For both of these two issues, data storage and data movement, the following points should be considered. safety When governanceData in motion and data at rest must be protected from unauthorized access, while making all that data available to analysts and tools that can mine the data for opportunities. Similarly, data governance can also be an issue. This is because moving data generated in one geographic location to a new location can raise government or tax issues.
Finally, a fourth factor to consider is analysisIt should be stored in a way that is accessible for analysis, updated frequently enough, properly cataloged and carefully organized.
A brief introduction to data modernization
Humans are curious creatures. What we create in real life is often recreated in the digital world. Many of us have cluttered homes and offices because we haven’t found the right place to store everything. Sadly, the same applies to how data is managed.
As I explained earlier, we’ve siled a lot of it. But even if you pull all that data into a central data lake, there’s no optimal way to search, sort, and sift through it all. Data modernization is about updating the way data is stored and retrieved to take advantage of the latest advances in big data, machine learning, AI, and even in-memory databases.
The IT buzzwords of data modernization and digital transformation go hand in hand. It has top data storage and retrieval methodologies (often of Top) Organizational IT priorities. This is called a data-first strategy, and it brings great benefits to your business.
Here it is. If data is trapped and trapped, it cannot be used effectively. Innovation is stifled if you and your team are constantly trying to find the data you need, or have never seen it in the first place. But freeing that data opens up new opportunities.
Not only that, poorly managed data can waste the time of dedicated IT staff. Instead of working to move the organization forward through innovation, they spend their time managing all these different systems, databases, and interfaces and troubleshooting everything that can break in different ways.
Modernizing your data not only means you can innovate, but it also means you free up time to think instead of react. It also gives you time to further develop applications and features that can open up new frontiers for your business.
Find hidden value and actionable insights in your data
The process of data modernization and adopting a data-first strategy can be daunting. Technologies such as cloud services and AI can help. Cloud services help by providing an on-demand, scalable infrastructure that can grow as the amount of data collected grows. AI empowers specialists and line-of-business managers to take action by providing tools that can sift through all data and organize it consistently.
However, it remains a major challenge for most IT teams. IT departments typically don’t silo all their data. It happens organically as more systems are installed and more of his to-do items are put on people’s lists.
That’s where management and infrastructure services like HPE GreenLake and its competitors can help. GreenLake offers a pay-as-you-go model, so you don’t have to “estimate” capacity usage upfront. With cross-application and cross-service dashboards and extensive expert support, HPE GreenLake helps you transform your data everywhere into a data-first strategy.
Leave a Reply