Post Syndicated from Vinodh Subramanian original https://backblazeprod.wpenginepowered.com/blog/whats-the-diff-bandwidth-vs-throughput/

You probably wouldn’t buy a car without knowing its horsepower. The metric might not matter as much to you as things like fuel efficiency, safety, or spiffy good looks. It might not even matter at all, but it’s still something you want to know before driving off the lot.

Similarly, you probably wouldn’t buy cloud storage without knowing a little bit about how it performs. Whether you need the metaphorical Ferrari of cloud providers, the safety features of a  Volvo, or the towing capacity of a semitruck, understanding how each performs can significantly impact your cloud storage decisions. And to understand cloud performance, you have to understand the difference between bandwidth and throughput.

In this blog, I’ll explain what bandwidth and throughput are and how they differ, as well as other key concepts like threading, multi-threading, and throttling—all of which can add more complexity and potential confusion to a cloud storage decision and the efficiency of data transfers. 

Bandwidth, Throughput, and Latency: A Primer

Three critical components form the cornerstone of cloud performance: bandwidth, throughput, and latency. To easily understand their impact, imagine the flow of data to water moving through a pipe—an analogy that paints a visual picture of how data travels across a network.

• Bandwidth: The diameter of the pipe represents bandwidth. It’s the maximum width that dictates how much water (data) can flow through it at any given time. In technical terms, bandwidth is the data transfer rate that a network connection can support. It’s usually measured in bits per second (bps). A wider pipe (higher bandwidth) means more data can flow, similar to having a multi-lane road where more vehicles can travel side by side.
• Throughput: If bandwidth is the pipe’s width, then throughput is the rate at which water moves through the pipe successfully. In the context of data, throughput is the actual data transfer rate that is sent over a network. It is also measured in bits per second (bps). Various factors can affect throughput—such as network traffic, processing power, packet loss, etc. While bandwidth is the potential capacity, throughput is the reality of performance, which is often less than the theoretical maximum due to real-world constraints. 
• Latency: Now, consider the time it takes for water to start flowing from the pipe’s opening after the tap is turned on. That time delay can be considered as latency. It’s the time it takes for a packet of data to travel from the source to the destination. Latency is crucial in use cases where time is of the essence, and even a slight delay can be detrimental to the user experience.

Understanding how bandwidth, throughput, and latency are interrelated is vital for anyone relying on cloud storage services. Bandwidth sets the stage for potential performance, but it’s the throughput that delivers actual results. Meanwhile, latency is a measure of how long it takes data to be delivered to the end user in real time. 

Threading and Multi-Threading in Cloud Storage

When we talk about moving data in the cloud, two concepts often come up: threading and multi-threading. These might sound very technical, but they’re actually pretty straightforward once broken down into simpler terms. 

First of all, threads go by many different names. Different applications may refer to them as streams, concurrent threads, parallel threads, concurrent uploads, parallelism, etc. But what all these terms refer to when we’re discussing cloud storage is the process of uploading files. To understand threads, think of a big pipe with a bunch of garden hoses running through it. The garden hose is a single thread in our pipe analogy.  The hose carries water (your data) from one point to another—say from your computer to the cloud or vice versa. In simple terms, it’s the pathway your data takes. Each hose represents an individual pathway through which data can move between a storage device and the network. 

Cloud storage systems use sophisticated algorithms to manage and prioritize threads. This ensures that resources are allocated efficiently to optimize data flow. Threads can be prioritized based on various criteria such as the type of data being transferred, network conditions, and overall load on the system.

Multi-Threading

Now, imagine: instead of just one garden hose within a pipe, you have several in parallel to each other. This setup is multi-threading. It lets multiple streams of water (data) flow at the same time, significantly speeding up the entire process. In the context of cloud storage, multi-threading enables the simultaneous transfer of multiple data streams, significantly speeding up data upload and download.

Cloud storage takes advantage of multithreading. It can take pretty much as many threads as you can throw at it and its performance should scale accordingly. But it doesn’t do so automatically—because the effectiveness of multi-threading depends on the underlying network infrastructure and the ability of the software to efficiently manage multiple threads. 

Chances are most devices can’t handle or take advantage of the maximum number of threads cloud storage can handle as it puts additional load on your network and device. Therefore, it often takes a trial-and-error approach to find the sweet spot to get optimal performance without severely affecting the usability of your device.

Managing Thread Count

Certain applications automatically manage threading and adjust the number of threads for optimal performance. When you’re using cloud storage with an integration like backup software or a network attached storage (NAS) device, the multi-threading setting is typically found in the integration’s settings. 

Many backup tools, like Veeam, are already set to multi-thread by default. However, some applications might default to using a single thread unless manually configured otherwise. 

That said, there are limitations associated with managing multiple threads. The gains from increasing the number of threads are limited by the bandwidth, processing power, and memory. Additionally, not all tasks are suitable for multi-threading; some processes need to be executed sequentially to maintain data integrity and dependencies between tasks. 

In essence, threading is about creating a pathway for your data and multi-threading is about creating multiple pathways to move more data at the same time. This makes storing and accessing files in the cloud much faster and more efficient. 

The Role of Throttling

Throttling is the deliberate slowing down of internet speed by service providers. In the pipe analogy, it’s similar to turning down the water flow from a faucet. Service providers use throttling to manage network traffic and prevent the system from becoming overloaded. By controlling the flow, they ensure that no single user or application monopolizes the bandwidth.

Why Do Cloud Service Providers Throttle?

The primary reason cloud service providers would throttle is to maintain an equitable distribution of network resources. During peak usage times, networks can become congested, much like roads during rush hour. Throttling helps manage these peak loads, ensuring all users have access to the network without significant drops in quality or service. It’s a balancing act, aiming to provide a steady, reliable service to as many users as possible. 

Scenarios Where Throttling Can Be a Hindrance

While throttling aims to manage network traffic for fairness purposes, it can be frustrating in certain situations. For heavy data users, such as businesses that rely on real-time data access and media teams uploading and downloading large files, throttling can slow operations and impact productivity. Additionally, for services not directly causing any congestion, throttling can seem unnecessary and restrictive. 

Optimizing Cloud Storage Performance

Achieving optimal performance in cloud storage involves more than just selecting a service; it requires a clear understanding of how bandwidth, throughput, latency, threading, and throttling interact and affect data transfer. Tailoring these elements to your specific needs can significantly enhance your cloud storage experience.

• Balancing bandwidth, throughput, and latency: The key to optimizing cloud performance lies in your use case. For real-time applications like video conferencing or gaming, low latency is crucial, whereas, for backup use cases, high throughput might be more important. Assessing the types of files you’re transferring and their size along with content delivery networks (CDN) can help in optimizing and achieving peak performance.
• Effective use of threading and multi-threading: Utilizing multi-threading effectively means understanding when it can be beneficial and when it might lead to diminishing returns. For large file transfers, multi-threading can significantly reduce transfer times. However, for smaller files, the overhead of managing multiple threads might outweigh the benefits. Using tools that automatically adjust the number of threads based on file size and network conditions can offer the best of both worlds.
• Navigating throttling for optimal performance: When selecting a cloud storage provider (CSP), it’s crucial to consider their throttling policies. Providers vary in how and when they throttle data transfer speeds, affecting performance. Understanding these policies upfront can help you choose a provider that aligns with your performance needs. 

In essence, optimizing cloud storage performance is an ongoing process of adjustment and adaptation. By carefully considering your specific needs, experimenting with settings, and staying informed about your provider’s policies, you can maximize the efficiency and effectiveness of your cloud storage solutions.

The post What’s the Diff: Bandwidth vs. Throughput appeared first on Backblaze Blog | Cloud Storage & Cloud Backup.

Post Syndicated from Molly Clancy original https://www.backblaze.com/blog/whats-the-diff-hot-and-cold-data-storage/

Temperature, specifically a range from cold to hot, is a common way to describe different levels of data storage. It’s possible these terms originated based on where data was historically stored. Hot data was stored close to the heat of the spinning drives and CPUs. Cold data was stored on drives or tape away from the warmer data center, likely tucked away on a shelf somewhere. 

Today, they’re used to describe how easily you can access your data. Hot storage is for data you need fast or access frequently. Cold storage is typically used for data you rarely need. The terms are used by most data storage providers to describe their tiered storage plans. However, there are no industry standard definitions for what hot and cold mean, which makes comparing services across different storage providers challenging. 

It’s a common misconception that hot storage means expensive storage and that cold storage means slower, less expensive storage. Today, we’ll explain why these terms may no longer be serving you when it comes to anticipating storage cost and performance.

Defining Hot Storage

Hot storage serves as the go-to destination for frequently accessed and mission-critical data that demands swift retrieval. Think of it as the fast lane of data storage, tailored for scenarios where time is of the essence. Industries relying on real-time data processing and rapid response times, such as video editing, web content, and application development, find hot storage to be indispensable.

To achieve the necessary rapid data access, hot storage is often housed in hybrid or tiered storage environments. The hotter the service, the more it embraces cutting-edge technologies, including the latest drives, fastest transport protocols, and geographical proximity to clients or multiple regions. However, the resource-intensive nature of hot storage warrants a premium, and leading cloud data storage providers like Microsoft’s Azure Hot Blobs and AWS S3 reflect this reality.

Data stored in the hottest tier might use solid-state drives (SSDs), which are optimized for lower latency and higher transactional rates compared to traditional hard drives. In other cases, hard disk drives are more suitable for environments where the drives are heavily accessed due to their higher durability standing up to intensive read/write cycles.

Regardless of the storage medium, hot data workloads necessitate fast and consistent response times, making them ideal for tasks like capturing telemetry data, messaging, and data transformation.

Defining Cold Storage

On the opposite end of the data storage spectrum lies cold storage, catering to information accessed infrequently and without the urgency of hot data. Cold storage houses data that might remain dormant for extended periods, months, years, decades, or maybe forever. Practical examples might include old projects or records mandated for financial, legal, HR, or other business record-keeping requirements.

Cold cloud storage systems prioritize durability and cost-effectiveness over real-time data manipulation capabilities. Services like Amazon Glacier and Google Coldline take this approach, offering slower retrieval and response times than their hot storage counterparts. Lower performing and less expensive storage environments, both on-premises and in the cloud, commonly host cold data. 

Linear Tape Open (LTO or Tape) has historically been a popular storage medium for cold data, though manual retrieval from storage racks renders it relatively slow. To access data from LTO, the tapes must be physically retrieved from storage racks and mounted in a tape reading machine, making it one of the slowest, therefore coldest, methods of storing data.

While cold cloud storage systems generally boast lower overall costs than warm or hot storage, they may incur higher per-operation expenses. Accessing data from cold storage demands patience and thoughtful planning, as the response times are intentionally sluggish.

With the landscape of data storage continually evolving, the definition of cold storage has also expanded. In modern contexts, cold storage might describe completely offline data storage, wherein information resides outside the cloud and remains disconnected from any network. This isolation, also described as air gapped, is crucial for safeguarding sensitive data. However, today, data can be virtually air-gapped using technology like Object Lock.

Traditional Views of Cold and Hot Data Storage

What Is Hot Cloud Storage?

Today there are new players in data storage, who, through innovation and efficiency, are able to offer cloud storage at the cost of cold storage, but with the performance and availability of hot storage.

The concept of organizing data by temperature has long been employed by diversified cloud providers like Amazon, Microsoft, and Google to describe their tiered storage services and set pricing accordingly. But, today, in a cloud landscape defined by the open, multi-cloud internet, customers have come to realize the value and benefits they can get from moving away from those diversified providers. 

A wave of independent cloud providers are disrupting the traditional notions of cloud storage temperatures, offering cloud storage that’s as cost-effective as cold storage, yet delivering the speed and availability associated with hot storage. If you’re familiar with Backblaze B2 Cloud Storage, you know where we’re going with this. 

Backblaze B2 falls into this category. We can compete on price with LTO and other traditionally cold storage services, but can be used for applications that are usually reserved for hot storage, such as media management, workflow collaboration, websites, and data retrieval.

The newfound efficiency of this model has prompted customers to rethink their storage strategies, opting to migrate entirely from cumbersome cold storage and archival systems.

What Temperature Is Your Cloud Storage?

When it comes to choosing the right storage temperature for your cloud data, organizations must carefully consider their unique needs. Ensuring that storage costs align with actual requirements is key to maintaining a healthy bottom line. The ongoing evolution of cloud storage services, driven by efficiency, technology, and innovation, further amplifies the need for tailored storage solutions.

Still have questions that aren’t answered here? Join the discussion in the comments.

The post What’s the Diff: Hot and Cold Data Storage appeared first on Backblaze Blog | Cloud Storage & Cloud Backup.

Post Syndicated from Molly Clancy original https://www.backblaze.com/blog/whats-the-diff-programs-processes-and-threads/

Programs, processes, and threads are all terms that relate to software execution, but you may not know what they really mean. Whether you’re a seasoned developer, an aspiring enthusiast, or you’re just wondering what you’re looking at when you open Task Manager on a PC or Activity Monitor on a Mac, learning these terms is essential for understanding how a computer works.

This post explains the technical concepts behind computer programs, processes, and threads to give you a better understanding of the functionality of your digital devices. With this knowledge, you can quickly diagnose problems and come up with solutions, like knowing if you need to install more memory for better performance. If you care about having a fast, efficient computer, it is worth taking the time to understand these key terms. 

What Is a Computer Program?

A program is a sequence of coded commands that tells a computer to perform a given task. There are many types of programs, including programs built into the operating system (OS) and ones to complete specific tasks. Generally, task-specific programs are called applications (or apps). For example, you are probably reading this post using a web browser application like Google Chrome, Mozilla Firefox, or Apple Safari. Other common applications include email clients, word processors, and games.

The process of creating a computer program involves designing algorithms, writing code in a programming language, and then compiling or interpreting that code to transform it into machine-readable instructions that the computer can execute.

Compiled vs. Interpreted Programs

Many programs are written in a compiled language and created using programming languages like C, C++, C#. The end result is a text file of code that is compiled into binary form in order to run on the computer (more on binary form in a few paragraphs). The text file speaks directly to your computer. While they’re typically fast, they are also fixed compared to interpreted programs. That has positives and negatives: you have more control over things like memory management, but you’re platform dependent and, if you have to change something in your code, it typically takes longer to build and test.

There is another kind of program called an interpreted program. They require an additional program to take your program instructions and translate that to code for your computer. Compared with compiled languages, these types of programs are platform-independent (you just have to find a different interpreter, instead of writing a whole new program) and they typically take up less space. Some of the most common interpreted programming languages are Python, PHP, JavaScript, and Ruby.

Ultimately, both kinds of programs are run and loaded into memory in binary form. Programs have to run in binary because your computer’s CPU understands only binary instructions.

What Is Binary Code?

Binary is the native language of computers. At their most basic level, computers use only two states of electrical current—on and off. The on state is represented by 1 and the off state is represented by 0. Binary is different from the number system—base 10—that we use in daily life. In base 10, each digit position can be anything from 0 to 9. In the binary system, also known as base 2, each position is either a 0 or a 1.

How Are Computer Programs Stored and Run?

Programs are typically stored on a disk or in nonvolatile memory in executable format. Let’s break that down to understand why.

In this context, we’ll talk about your computer having two types of memory: volatile and nonvolatile. Volatile memory is temporary and processes in real time. It’s faster, easily accessible, and increases the efficiency of your computer. However, it’s not permanent. When your computer turns off, this type of memory resets.

Nonvolatile memory, on the other hand, is permanent unless deleted. While it’s slower to access, it can store more information. So, that makes it a better place to store programs. A file in an executable format is simply one that runs a program. It can be run directly by your CPU (that’s your processor). Examples of these file types are .exe in Windows and .app in Mac.

What Resources Does a Program Need to Run?

Once a program has been loaded into memory in binary form, what happens next?

Your executing program needs resources from the OS and memory to run. Without these resources, you can’t use the program. Fortunately, your OS manages the work of allocating resources to your programs automatically. Whether you use Microsoft Windows, macOS, Linux, Android, or something else, your OS is always hard at work directing your computer’s resources needed to turn your program into a running process.

In addition to OS and memory resources, there are a few essential resources that every program needs.

• Register. Think of a register as a holding pen that contains data that may be needed by a process like instructions, storage addresses, or other data.
• Program counter. Also known as an instruction pointer, the program counter plays an organizational role. It keeps track of where a computer is in its program sequence.
• Stack. A stack is a data structure that stores information about the active subroutines of a computer program. It is used as scratch space for the process. It is distinguished from dynamically allocated memory for the process that is known as the “heap.”

What Is a Computer Process?

When a program is loaded into memory along with all the resources it needs to operate, it is called a process. You might have multiple instances of a single program. In that situation, each instance of that running program is a process. 

Each process has a separate memory address space. That separate memory address is helpful because it means that a process runs independently and is isolated from other processes. However, processes cannot directly access shared data in other processes. Switching from one process to another requires some amount of time (relatively speaking) for saving and loading registers, memory maps, and other resources.

Having independent processes matters for users because it means one process won’t corrupt or wreak havoc on other processes. If a single process has a problem, you can close that program and keep using your computer. Practically, that means you can end a malfunctioning program and keep working with minimal disruptions.

What Are Threads?

The final piece of the puzzle is threads. A thread is the unit of execution within a process.

When a process starts, it receives an assignment of memory and other computing resources. Each thread in the process shares that memory and resources. With single-threaded processes, the process contains one thread.

In multi-threaded processes, the process contains more than one thread, and the process is accomplishing a number of things at the same time (to be more accurate, we should say “virtually” the same time—you can read more about that in the section below on concurrency).

Earlier, we talked about the stack and the heap, the two kinds of memory available to a thread or process. Distinguishing between these kinds of memory matters because each thread will have its own stack. However, all the threads in a process will share the heap.

Some people call threads lightweight processes because they have their own stack but can access shared data. Since threads share the same address space as the process and other threads within the process, it is easy to communicate between the threads. The disadvantage is that one malfunctioning thread in a process can impact the viability of the process itself.

How Threads and Processes Work Step By Step

Here’s what happens when you open an application on your computer.

• The program starts out as a text file of programming code.
• The program is compiled or interpreted into binary form.
• The program is loaded into memory.
• The program becomes one or more running processes. Processes are typically independent of one another.
• Threads exist as the subset of a process.
• Threads can communicate with each other more easily than processes can.
• Threads are more vulnerable to problems caused by other threads in the same process.

Computer Process vs. Threads

What About Concurrency and Parallelism?

A question you might ask is whether processes or threads can run at the same time. The answer is: it depends. In environments with multiple processors or CPU cores, simultaneous execution of multiple processes or threads is feasible. However, on a single processor system, true simultaneous execution isn’t possible. In these cases, a process scheduling algorithm is employed to share the CPU among running processes or threads, creating the illusion of parallel execution. Each task is allocated a “time slice,” and the swift switching between tasks occurs seamlessly, typically imperceptible to users. The terms “parallelism” (denoting genuine simultaneous execution) and “concurrency” (indicating the interleaving of processes over time to simulate simultaneous execution) distinguish between the two modes of operation, whether truly simultaneous or approximated.

How Google Chrome Uses Processes and Threads

To illustrate the impact of processes and threads, let’s consider a real-world example with a program that many of us use, Google Chrome. 

When Google designed the Chrome browser, they faced several important decisions. For instance, how should Chrome handle the fact that many different tasks often happen at the same time when using a browser? Every browser window (or tab) may communicate with several servers on the internet to download audio, video, text, and other resources. In addition, many users have 10 to 20 browser tabs (or more…) open most of the time, and each of these tabs may perform multiple tasks.

Google had to decide how to handle all of these tasks. They chose to run each browser window in Chrome as a separate process rather than a thread or many threads. That approach brought several benefits.

• Running each window as a process protects the overall application from bugs and glitches.
• Isolating a JavaScript program in a process prevents it from using too much CPU time and memory and making the entire browser unresponsive.

That said, there is a trade-off cost to Google’s design decision. Starting a new process for each browser window has a higher fixed cost in memory and resources compared to using threads. They were betting that their approach would end up with less memory bloat overall.

Using processes instead of threads provides better memory usage when memory is low. In practice, an inactive browser window is treated as a lower priority. That means the operating system may swap it to disk when memory is needed for other processes. If the windows were threaded, it would be more difficult to allocate memory efficiently which ultimately leads to lost computer performance.

For more insights on Google’s design decisions for Chrome on Google’s Chromium Blog or on the Chrome Introduction Comic.

The screen capture below shows the Google Chrome processes running on a MacBook Air with many tabs open. You can see that some Chrome processes are using a fair amount of CPU time and resources (e.g., the one at the top is using 44 threads) while others are using fewer.

The Activity Monitor on the Mac (or Task Manager in Windows) on your system can be a valuable ally in fine-tuning your computer or troubleshooting problems. If your computer is running slowly or a program or browser window isn’t responding for a while, you can check its status using the system monitor.

In some cases, you’ll see a process marked as “Not Responding.” Try quitting that process and see if your system runs better. If an application is a memory hog, you might consider choosing a different application that will accomplish the same task.

Made It This Far?

We hope this Tron-like dive into the fascinating world of computer programs, processes, and threads has cleared up some questions.

At the start, we promised clarity on using these terms to improve performance. You can use Activity Monitor on the Mac or Task Manager on Windows to close applications and processes that are malfunctioning. That’s beneficial because it means you can end a malfunctioning program without the hassle of turning off your computer.

Still have questions? We’d love to hear from you in the comments.

FAQ

The post What’s the Diff: Programs, Processes, and Threads appeared first on Backblaze Blog | Cloud Storage & Cloud Backup.