|By Brendan Gregg||
|January 14, 2014 11:00 AM EST||
"This excerpt is from the book, "Systems Performance: Enterprise and the Cloud", authored by Brendan Gregg, published by Prentice Hall Professional, Oct. 2013, ISBN 9780133390094, Copyright © 2014 Pearson Education, Inc. For more info, please visit the publisher site:
CPUs drive all software and are often the first target for systems performance analysis. Modern systems typically have many CPUs, which are shared among all running software by the kernel scheduler. When there is more demand for CPU resources than there are resources available, process threads (or tasks) will queue, waiting their turn. Waiting can add significant latency during the runtime of applications, degrading performance.
The usage of the CPUs can be examined in detail to look for performance improvements, including eliminating unnecessary work. At a high level, CPU usage by process, thread, or task can be examined. At a lower level, the code path within applications and the kernel can be profiled and studied. At the lowest level, CPU instruction execution and cycle behavior can be studied.
This chapter consists of five parts:
- Background introduces CPU-related terminology, basic models of CPUs, and key CPU performance concepts.
- Architecture introduces processor and kernel scheduler architecture.
- Methodology describes performance analysis methodologies, both observa- tional and experimental.
- Analysis describes CPU performance analysis tools on Linux- and Solaris- based systems, including profiling, tracing, and visualizations.
- Tuning includes examples of tunable parameters.
The first three sections provide the basis for CPU analysis, and the last two show its practical application to Linux- and Solaris-based systems.
The effects of memory I/O on CPU performance are covered, including CPU cycles stalled on memory and the performance of CPU caches. Chapter 7, Memory, continues the discussion of memory I/O, including MMU, NUMA/UMA, system interconnects, and memory busses.
For reference, CPU-related terminology used in this chapter includes the following:
- Processor: the physical chip that plugs into a socket on the system or pro- cessor board and contains one or more CPUs implemented as cores or hard- ware threads.
- Core: an independent CPU instance on a multicore processor. The use of cores is a way to scale processors, called chip-level multiprocessing (CMP).
- Hardware thread: a CPU architecture that supports executing multiple threads in parallel on a single core (including Intel's Hyper-Threading Tech- nology), where each thread is an independent CPU instance. One name for this scaling approach is multithreading.
- CPU instruction: a single CPU operation, from its instruction set. There are instructions for arithmetic operations, memory I/O, and control logic.
- Logical CPU: also called a virtual processor,1 an operating system CPU instance (a schedulable CPU entity). This may be implemented by the processor as a hardware thread (in which case it may also be called a virtual core), a core, or a single-core processor.
- Scheduler: the kernel subsystem that assigns threads to run on CPUs.
- Run queue: a queue of runnable threads that are waiting to be serviced by
- CPUs. For Solaris, it is often called a dispatcher queue.
Other terms are introduced throughout this chapter. The Glossary includes basic terminology for reference, including CPU, CPU cycle, and stack. Also see the terminology sections in Chapters 2 and 3.
The following simple models illustrate some basic principles of CPUs and CPU per- formance. Section 6.4, Architecture, digs much deeper and includes implementation- specific details.
Figure 1 shows an example CPU architecture, for a single processor with four cores and eight hardware threads in total. The physical architecture is pictured, along with how it is seen by the operating system.
Figure 1: CPU architecture
Each hardware thread is addressable as a logical CPU, so this processor appears as eight CPUs. The operating system may have some additional knowledge of topology, such as which CPUs are on the same core, to improve its scheduling decisions.
CPU Memory Caches
Processors provide various hardware caches for improving memory I/O perfor- mance. Figure 2 shows the relationship of cache sizes, which become smaller and faster (a trade-off) the closer they are to the CPU.
The caches that are present, and whether they are on the processor (integrated) or external to the processor, depend on the processor type. Earlier processors pro- vided fewer levels of integrated cache.
Figure 2: CPU cache sizes
CPU Run Queues
Figure 3 shows a CPU run queue, which is managed by the kernel scheduler.
Figure 3: CPU run queue
The thread states shown in the figure, ready to run and on-CPU, are covered in Figure 3.7 in Chapter 3, Operating Systems.
The number of software threads that are queued and ready to run is an impor- tant performance metric indicating CPU saturation. In this figure (at this instant) there are four, with an additional thread running on-CPU. The time spent waiting on a CPU run queue is sometimes called run-queue latency or dispatcher-queue latency. In this book, the term scheduler latency is used instead, as it is appropri- ate for all dispatcher types, including those that do not use queues (see the discus- sion of CFS in Section 6.4.2, Software).
For multiprocessor systems, the kernel typically provides a run queue for each CPU and aims to keep threads on the same run queue. This means that threads are more likely to keep running on the same CPUs, where the CPU caches have cached their data. (These caches are described as having cache warmth, and the approach to favor CPUs is called CPU affinity.) On NUMA systems, memory locality may also be improved, which also improves performance (this is described in Chapter 7, Memory).
It also avoids the cost of thread synchronization (mutex locks) for queue operations, which would hurt scalability if the run queue was global and shared among all CPUs.
The following are a selection of important concepts regarding CPU performance, beginning with a summary of processor internals: the CPU clock rate and how instructions are executed. This is background for later performance analysis, particularly for understanding the cycles-per-instruction (CPI) metric.
The clock is a digital signal that drives all processor logic. Each CPU instruction may take one or more cycles of the clock (called CPU cycles) to execute. CPUs exe- cute at a particular clock rate; for example, a 5 GHz CPU performs 5 billion clock cycles per second.
Some processors are able to vary their clock rate, increasing it to improve performance or decreasing it to reduce power consumption. The rate may be varied on request by the operating system, or dynamically by the processor itself. The ker- nel idle thread, for example, can request the CPU to throttle down to save power.
Clock rate is often marketed as the primary feature of the processor, but this can be a little misleading. Even if the CPU in your system appears to be fully utilized (a bottleneck), a faster clock rate may not speed up performance-it depends on what those fast CPU cycles are actually doing. If they are mostly stall cycles while waiting on memory access, executing them more quickly doesn't actually increase the CPU instruction rate or workload throughput.
CPUs execute instructions chosen from their instruction set. An instruction includes the following steps, each processed by a component of the CPU called a functional unit:
- Instruction fetch
- Instruction decode
- Memory access
- Register write-back
The last two steps are optional, depending on the instruction. Many instructions operate on registers only and do not require the memory access step.
Each of these steps takes at least a single clock cycle to be executed. Memory access is often the slowest, as it may take dozens of clock cycles to read or write to main memory, during which instruction execution has stalled (and these cycles while stalled are called stall cycles). This is why CPU caching is important, as described in Section 6.4: it can dramatically reduce the number of cycles needed for memory access.
The instruction pipeline is a CPU architecture that can execute multiple instructions in parallel, by executing different components of different instructions at the same time. It is similar to a factory assembly line, where stages of production can be executed in parallel, increasing throughput.
Consider the instruction steps previously listed. If each were to take a single clock cycle, it would take five cycles to complete the instruction. At each step of this instruction, only one functional unit is active and four are idle. By use of pipe- lining, multiple functional units can be active at the same time, processing differ- ent instructions in the pipeline. Ideally, the processor can then complete one instruction with every clock cycle.
But we can go faster still. Multiple functional units can be included of the same type, so that even more instructions can make forward progress with each clock cycle. This CPU architecture is called superscalar and is typically used with pipe- lining to achieve a high instruction throughput.
The instruction width describes the target number of instructions to process in parallel. Modern processors are 3-wide or 4-wide, meaning they can complete up to three or four instructions per cycle. How this works depends on the processor, as there may be different numbers of functional units for each stage.
Cycles per instruction (CPI) is an important high-level metric for describing where a CPU is spending its clock cycles and for understanding the nature of CPU utilization. This metric may also be expressed as instructions per cycle (IPC), the inverse of CPI.
A high CPI indicates that CPUs are often stalled, typically for memory access. A low CPI indicates that CPUs are often not stalled and have a high instruction throughput. These metrics suggest where performance tuning efforts may be best spent.
Memory-intensive workloads, for example, may be improved by installing faster memory (DRAM), improving memory locality (software configuration), or reducing the amount of memory I/O. Installing CPUs with a higher clock rate may not improve performance to the degree expected, as the CPUs may need to wait the same amount of time for memory I/O to complete. Put differently, a faster CPU may mean more stall cycles but the same rate of completed instructions.
The actual values for high or low CPI are dependent on the processor and processor features and can be determined experimentally by running known work- loads. As an example, you may find that high-CPI workloads run with a CPI at ten or higher, and low CPI workloads run with a CPI at less than one (which is possi- ble due to instruction pipelining and width, described earlier).
It should be noted that CPI shows the efficiency of instruction processing, but not of the instructions themselves. Consider a software change that added an inefficient software loop, which operates mostly on CPU registers (no stall cycles): such a change may result in a lower overall CPI, but higher CPU usage and utilization.
CPU utilization is measured by the time a CPU instance is busy performing work during an interval, expressed as a percentage. It can be measured as the time a CPU is not running the kernel idle thread but is instead running user-level application threads or other kernel threads, or processing interrupts.
High CPU utilization may not necessarily be a problem, but rather a sign that the system is doing work. Some people also consider this an ROI indicator: a highly utilized system is considered to have good ROI, whereas an idle system is considered wasted. Unlike with other resource types (disks), performance does not degrade steeply under high utilization, as the kernel supports priorities, preemption, and time sharing. These together allow the kernel to understand what has higher priority, and to ensure that it runs first.
The measure of CPU utilization spans all clock cycles for eligible activities, including memory stall cycles. It may seem a little counterintuitive, but a CPU may be highly utilized because it is often stalled waiting for memory I/O, not just executing instructions, as described in the previous section.
CPU utilization is often split into separate kernel- and user-time metrics.
SYS-CON Events announced today that G2G3 will exhibit at SYS-CON's @DevOpsSummit Silicon Valley, which will take place on November 3–5, 2015, at the Santa Clara Convention Center in Santa Clara, CA. Based on a collective appreciation for user experience, design, and technology, G2G3 is uniquely qualified and motivated to redefine how organizations and people engage in an increasingly digital world.
Oct. 6, 2015 08:15 PM EDT Reads: 337
If you are new to Python, you might be confused about the different versions that are available. Although Python 3 is the latest generation of the language, many programmers still use Python 2.7, the final update to Python 2, which was released in 2010. There is currently no clear-cut answer to the question of which version of Python you should use; the decision depends on what you want to achieve. While Python 3 is clearly the future of the language, some programmers choose to remain with Py...
Oct. 6, 2015 08:00 PM EDT Reads: 177
“All our customers are looking at the cloud ecosystem as an important part of their overall product strategy. Some see it evolve as a multi-cloud / hybrid cloud strategy, while others are embracing all forms of cloud offerings like PaaS, IaaS and SaaS in their solutions,” noted Suhas Joshi, Vice President – Technology, at Harbinger Group, in this exclusive Q&A with Cloud Expo Conference Chair Roger Strukhoff.
Oct. 6, 2015 02:45 PM EDT Reads: 370
Opinions on how best to package and deliver applications are legion and, like many other aspects of the software world, are subject to recurring trend cycles. On the server-side, the current favorite is container delivery: a “full stack” approach in which your application and everything it needs to run are specified in a container definition. That definition is then “compiled” down to a container image and deployed by retrieving the image and passing it to a container runtime to create a running...
Oct. 6, 2015 12:45 PM EDT Reads: 130
Clearly the way forward is to move to cloud be it bare metal, VMs or containers. One aspect of the current public clouds that is slowing this cloud migration is cloud lock-in. Every cloud vendor is trying to make it very difficult to move out once a customer has chosen their cloud. In his session at 17th Cloud Expo, Naveen Nimmu, CEO of Clouber, Inc., will advocate that making the inter-cloud migration as simple as changing airlines would help the entire industry to quickly adopt the cloud wit...
Oct. 6, 2015 12:30 PM EDT Reads: 581
As the world moves towards more DevOps and microservices, application deployment to the cloud ought to become a lot simpler. 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 Cloud Foundry - a complete developer-oriented Platform as a Service (PaaS) that is IaaS agnostic and supports vCloud, OpenStack and AWS. In his session at 17th Cloud Expo, Raghavan "Rags" Srinivas, an Architect/Developer Evangeli...
Oct. 6, 2015 12:15 PM EDT Reads: 106
Culture is the most important ingredient of DevOps. The challenge for most organizations is defining and communicating a vision of beneficial DevOps culture for their organizations, and then facilitating the changes needed to achieve that. Often this comes down to an ability to provide true leadership. As a CIO, are your direct reports IT managers or are they IT leaders? The hard truth is that many IT managers have risen through the ranks based on their technical skills, not their leadership ab...
Oct. 6, 2015 11:00 AM EDT Reads: 850
Apps and devices shouldn't stop working when there's limited or no network connectivity. Learn how to bring data stored in a cloud database to the edge of the network (and back again) whenever an Internet connection is available. In his session at 17th Cloud Expo, Bradley Holt, Developer Advocate at IBM Cloud Data Services, will demonstrate techniques for replicating cloud databases with devices in order to build offline-first mobile or Internet of Things (IoT) apps that can provide a better, ...
Oct. 6, 2015 10:45 AM EDT Reads: 452
Despite all the talk about public cloud services and DevOps, you would think the move to cloud for enterprises is clear and simple. But in a survey of almost 1,600 IT decision makers across the USA and Europe, the state of the cloud in enterprise today is still fraught with considerable frustration. The business case for apps in the real world cloud is hybrid, bimodal, multi-platform, and difficult. Download this report commissioned by NTT Communications to see the insightful findings – registra...
Oct. 6, 2015 10:00 AM EDT Reads: 212
Application availability is not just the measure of “being up”. Many apps can claim that status. Technically they are running and responding to requests, but at a rate which users would certainly interpret as being down. That’s because excessive load times can (and will be) interpreted as “not available.” That’s why it’s important to view ensuring application availability as requiring attention to all its composite parts: scalability, performance, and security.
Oct. 6, 2015 09:00 AM EDT Reads: 349
SYS-CON Events announced today that HPM Networks will exhibit at the 17th International Cloud Expo®, which will take place on November 3–5, 2015, at the Santa Clara Convention Center in Santa Clara, CA. For 20 years, HPM Networks has been integrating technology solutions that solve complex business challenges. HPM Networks has designed solutions for both SMB and enterprise customers throughout the San Francisco Bay Area.
Oct. 6, 2015 09:00 AM EDT Reads: 568
There once was a time when testers operated on their own, in isolation. They’d huddle as a group around the harsh glow of dozens of CRT monitors, clicking through GUIs and recording results. Anxiously, they’d wait for the developers in the other room to fix the bugs they found, yet they’d frequently leave the office disappointed as issues were filed away as non-critical. These teams would rarely interact, save for those scarce moments when a coder would wander in needing to reproduce a particula...
Oct. 6, 2015 08:45 AM EDT Reads: 258
What Is Emergent About Emergent Architecture? By @TheEbizWizard | @DevOpsSummit #DevOps #BigData #API
All we need to do is have our teams self-organize, and behold! Emergent design and/or architecture springs up out of the nothingness! If only it were that easy, right? I follow in the footsteps of so many people who have long wondered at the meanings of such simple words, as though they were dogma from on high. Emerge? Self-organizing? Profound, to be sure. But what do we really make of this sentence?
Oct. 6, 2015 08:00 AM EDT Reads: 379
As we increasingly rely on technology to improve the quality and efficiency of our personal and professional lives, software has become the key business differentiator. Organizations must release software faster, as well as ensure the safety, security, and reliability of their applications. The option to make trade-offs between time and quality no longer exists—software teams must deliver quality and speed. To meet these expectations, businesses have shifted from more traditional approaches of d...
Oct. 6, 2015 07:45 AM EDT Reads: 155
Information overload has infiltrated our lives. From the amount of news available and at our fingertips 24/7, to the endless choices we have when making a simple purchase, to the quantity of emails we receive on a given day, it’s increasingly difficult to sift out the details that really matter. When you envision your cloud monitoring system, the same thinking applies. We receive a lot of useless data that gets fed into the system, and the reality is no one in IT or DevOps has the time to manu...
Oct. 6, 2015 07:00 AM EDT Reads: 494
Last month, my partners in crime – Carmen DeArdo from Nationwide, Lee Reid, my colleague from IBM and I wrote a 3-part series of blog posts on DevOps.com. We titled our posts the Simple Math, Calculus and Art of DevOps. I would venture to say these are must-reads for any organization adopting DevOps. We examined all three ascpects – the Cultural, Automation and Process improvement side of DevOps. One of the key underlying themes of the three posts was the need for Cultural change – things like t...
Oct. 6, 2015 04:15 AM EDT Reads: 274
It is with great pleasure that I am able to announce that Jesse Proudman, Blue Box CTO, has been appointed to the position of IBM Distinguished Engineer. Jesse is the first employee at Blue Box to receive this honor, and I’m quite confident there will be more to follow given the amazing talent at Blue Box with whom I have had the pleasure to collaborate. I’d like to provide an overview of what it means to become an IBM Distinguished Engineer.
Oct. 6, 2015 03:45 AM EDT Reads: 135
I’ve been thinking a bit about microservices (μServices) recently. My immediate reaction is to think: “Isn’t this just yet another new term for the same stuff, Web Services->SOA->APIs->Microservices?” Followed shortly by the thought, “well yes it is, but there are some important differences/distinguishing factors.” Microservices is an evolutionary paradigm born out of the need for simplicity (i.e., get away from the ESB) and alignment with agile (think DevOps) and scalable (think Containerizati...
Oct. 6, 2015 03:00 AM EDT Reads: 2,460
The cloud has reached mainstream IT. Those 18.7 million data centers out there (server closets to corporate data centers to colocation deployments) are moving to the cloud. In his session at 17th Cloud Expo, Achim Weiss, CEO & co-founder of ProfitBricks, will share how two companies – one in the U.S. and one in Germany – are achieving their goals with cloud infrastructure. More than a case study, he will share the details of how they prioritized their cloud computing infrastructure deployments ...
Oct. 6, 2015 03:00 AM EDT Reads: 678
DevOps Summit at Cloud Expo 2014 Silicon Valley was a terrific event for us. The Qubell booth was crowded on all three days. We ran demos every 30 minutes with folks lining up to get a seat and usually standing around. It was great to meet and talk to over 500 people! My keynote was well received and so was Stan's joint presentation with RingCentral on Devops for BigData. I also participated in two Power Panels – ‘Women in Technology’ and ‘Why DevOps Is Even More Important than You Think,’ both ...
Oct. 6, 2015 02:00 AM EDT Reads: 8,588