Index
A
- access control, User Management and Access Control
- ad hoc configuration, Ad Hockery
- aggregation servers, Fundamentals and Elements of Metric Collection Systems
- agility, maintaining in presence of conflicting requests and changing priorities, Make System Stats Tell Stories
- Amazon CloudWatch, Autoscaling on Amazon EC2
- Amazon EC2, Load Balancing
- Amazon Elastic Block Store (EBS), Automated Installation Tools
- Amazon machine images (AMIs), Automated Installation Tools
- Amazon Web Services (AWS), Background
- Aminator tool, Automated Installation Tools
- APIs
- application-level metrics, Application-Level Measurement
- architecture-specific capacity planning, Predicting When Systems Will Fail
- architectures
- database, federated or sharded, Tying Application Level Metrics to System Statistics: Database Example
- decisions on, Architecture Decisions-Disaster Recovery
- design resources, Architecture Decisions
- disaster recovery, Disaster Recovery-Disaster Recovery
- hardware (vertical, horizontal, and diagonal scaling), Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)-Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)
- providing measurement points, Providing Measurement Points-Providing Measurement Points
- resource ceilings, Resource Ceilings
- effect on capacity, The Effects of Social Websites and Open APIs
- installation, configuration, and management architecture, Goal 2: All Changes Happen in One Place
- SLAs effect on architectural design, SLAs
- artificial load testing, limitations of, Finding web server ceilings in a load-balancing environment
- ATA/SATA protocol hard disks, Storage Capacity
- attention span of humans, Preliminaries
- ATTO Disk Benchmark, Storage Capacity
- authentication and authorization, User Management and Access Control
- Autobench, Finding web server ceilings in a load-balancing environment
- automated configuration (see configuration)
- automated deployment philosophies, Automated Deployment Philosophies-Goal 5: Maintain Consistency for Easier Troubleshooting
- goal 1, minimizing time to provision new capacity, Goal 1: Minimize Time to Provision New Capacity
- goal 2, all changes happen in one place, Goal 2: All Changes Happen in One Place
- goal 3, never log in ot individual server (for management), Goal 3: Never Log in to an Individual Server (for Management)
- goal 4, having new servers start working automatically, Goal 4: Have New Servers Start Working Automatically
- goal 5, maintaining consistency for easier toubleshooting, Goal 5: Maintain Consistency for Easier Troubleshooting
- automated installation and configuration tools, Deployment
- automated installation tools, Automated Installation Tools-The Installation Process
- automating forecasting, Automating the Forecasting-Automating the Forecasting
- autoscaling, Quick and Dirty Math, Autoscaling-Resources
- advanced approaches, Advanced Approaches
- by fixed amount, Autoscaling by Fixed Amount-Scaling by Percentage
- challenges of, The Challenge
- design guidelines, Design Guidelines
- miscellaneous approaches, Potpourri-Potpourri
- on Amazon EC2, Autoscaling on Amazon EC2-Autoscaling on Amazon EC2
- scalability analysis, Scalability Analysis
- scale-up events, properties of, Properties
- Spot instances, market in public clouds, The Challenge
- startup time aware, Startup Time Aware Scaling-Potpourri
- availability
- Availability Zones (AZs), Autoscaling
- AWS (see Amazon Web Services)
C
- cache eviction, Cache efficiency: working sets and dynamic data
- caching, Cache but Serve Stale
- caching systems, Caching Systems-Cache efficiency: working sets and dynamic data
- Cacti, Capacity Tracking Tools
- capacity
- capacity headroom
- capacity planning
- capacity tools, Capacity Tools-Books on Queuing Theory and the Mathematics of Capacity Planning
- capital expenditure (capex)
- ceilings (resource), Resource Ceilings
- cfityk curve-fitting GUI tool, Automating the Forecasting-Automating the Forecasting
- client/server relationship, artificial load testing and, Finding web server ceilings in a load-balancing environment
- cloud
- adding capacity, Buying Stuff
- elasticity of the cloud, making it easier not to under-provision, Quick and Dirty Math
- keeping close tabs on expected operational expenditures for, Make System Stats Tell Stories
- metrics collection in, Measurement: Units of Capacity
- multitenant environment, capacity measurements in, Special Use and Multiple Use Servers
- outages in, Summary
- performance variability in, Overview
- use of public clouds, Background
- workload analysis in, Performance and Capacity: Two Different Animals
- cloud jail, Buying Stuff
- CloudWatch (see Amazon CloudWatch)
- cluster management tools, Cluster Management/Container Orchestration
- cluster underutilization, Finding web server ceilings in a load-balancing environment
- coefficient of determination, Trends, Curves, and Time, Trends, Curves, and Time
- collectd, Fundamentals and Elements of Metric Collection Systems
- complexities of applications, growth of, Make System Stats Tell Stories
- configuration
- configuration tools, automated, Deployment, Automated Configuration
- consistency, guaranteeing, External Service Monitoring
- consumption rates (storage capacity), Consumption rates
- consumption-driven metrics, Trends, Curves, and Time
- containerization, Background
- containers, Overview
- content delivery networks (CDNs), Cache but Serve Stale
- continuous delivery (CD) platforms, Buying Stuff
- costs
- Cox, J., Make System Stats Tell Stories, Make System Stats Tell Stories
- CPU usage
- curve fitting, Trends, Curves, and Time-Trends, Curves, and Time
D
- daemons, for metrics collection, Fundamentals and Elements of Metric Collection Systems
- data sets, small, caveats concerning, Caveats Concerning Small Datasets
- database capacity, Database Capacity-Finding database ceilings
- database servers
- databases
- datacenters, Background
- in-house, cost of public cloud services versus, Autoscaling
- large companies with, procurement process, Buying Stuff
- migration from public cloud to, Quick and Dirty Math
- multiple, forecasting considerations for, Traffic Pattern Changes
- outages, Summary
- projected growth of spending on, The Effects of Social Websites and Open APIs
- using automated configuration for multiple datacenters, Example 2: Multiple Datacenters
- virtualization in, Looking Back and Moving forward
- Debian FAI, Automated Installation Tools
- dependencies between components, causing bottlenecks, Make System Stats Tell Stories
- dependent events (in capacity planning), Make System Stats Tell Stories
- deployment, Just-in-Time Inventory, Deployment-Readings
- automated configuration, Automated Configuration-Example 2: Multiple Datacenters
- automated deployment philosophies, Automated Deployment Philosophies-Goal 5: Maintain Consistency for Easier Troubleshooting
- goal 1, mimimizing time to provision new capacity, Goal 1: Minimize Time to Provision New Capacity
- goal 2, all changes happen in one place, Goal 2: All Changes Happen in One Place
- goal 3, never log into individual server (for management), Goal 3: Never Log in to an Individual Server (for Management)
- goal 4, having new servers start working automatically, Goal 4: Have New Servers Start Working Automatically
- goal 5, maintaining consistency for easier troubleshooting, Goal 5: Maintain Consistency for Easier Troubleshooting
- automated installation tools, Automated Installation Tools-The Installation Process
- deploying new equipment, Just-in-Time Inventory
- importance in capacity planning, Make System Stats Tell Stories
- inventory management as part of structure, Automated Configuration
- tools for, Deployment Tools
- DHCP (Dynamic Host Configuration Protocol), The Installation Process
- diagonal scaling, Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)
- disabling heavy features, Graceful Degradation and Disabling Heavy Features
- disaster recovery, Disaster Recovery-Disaster Recovery
- disk-image approach to imanging new machines, Automated Installation Tools
- Docker, Background
- downtime
- Dynamic Host Configuration Protocol (DHCP), The Installation Process
F
- factor of safety, Safety Factors
- failure
- features (heavy), disabling, Graceful Degradation and Disabling Heavy Features
- federated (or sharded) database architecture, Tying Application Level Metrics to System Statistics: Database Example
- field-programmable gate arrays (FPGAs), instance types with, Goal 5: Maintain Consistency for Easier Troubleshooting, Looking Back and Moving forward
- financial data, correlating with application and system metrics, Application-Level Measurement
- fityk curve-fitting program, Automating the Forecasting
- five-nines (SLAs and availability), SLAs
- fixed number of instances, autoscaling by, Autoscaling by Fixed Amount-Scaling by Percentage
- forecasting, Trends, Curves, and Time
- Fowler, Martin, Background
G
- Ganglia, Capacity Tracking Tools, Ganglia, Metric Collection and Event Notification Systems
- goals
- gold client image, Automated Installation Tools
- Goldratt, E.M., Make System Stats Tell Stories, Make System Stats Tell Stories
- Google Preemptible VMs, The Challenge
- graceful degradation, Graceful Degradation and Disabling Heavy Features
- graphics processing units (GPUs), Goal 5: Maintain Consistency for Easier Troubleshooting, Looking Back and Moving forward
- graphing tools, Ad Hoc Measurement and Graphing Tools
- Graphite, Capacity Tracking Tools
- growth
H
- hard disk drives, Storage Capacity
- Hard Disk Sentinel tool, Storage Capacity
- hardware
- becoming cheaper and faster, Predicting Trends
- centralized configuration and management environment, Goal 2: All Changes Happen in One Place
- decisions on, vertical, horizontal, and diagonal scaling, Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)-Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)
- diversity of hardware on public clouds, Overview
- homogenizing to prevent headaches, Goal 5: Maintain Consistency for Easier Troubleshooting
- inventory management for, Automated Configuration
- not buying enough versus too much, Preliminaries
- resource ceilings, Resource Ceilings
- hardware VM (HVM), Automated Installation Tools
- HDDs (see hard disk drives)
- high availability, SLAs
- historical data, Diagonal Scaling Opportunities
- horizontal scaling, Quick and Dirty Math, Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)
- HTTP server service, Defining Roles and Services
- Httperf, Finding web server ceilings in a load-balancing environment, A real-world example: cache measurement
- human attention span, Preliminaries
- HVM (hardware VM) AMIs, Automated Installation Tools
I
- iClassify, Automated Configuration
- imaging new machines, Automated Installation Tools
- infrastructure as code, Deployment
- infrastructure monitoring services, Capacity Tracking Tools
- infrastructure, assessing current working of, Preliminaries
- innovation, increasing rate of, Make System Stats Tell Stories
- Instagram, user base and traffic, Business Capacity Requirements
- installation tools, automated, Deployment, Automated Installation Tools-The Installation Process, Automated OS Installation
- instance types
- intent of users, External Service Monitoring
- interaction of users with content, External Service Monitoring
- inventory management, Automated Configuration
- IOBench, Storage Capacity
L
- LAMP (Linux, Apache, MySQL, and PHP) stack, Predicting When Systems Will Fail, Round-Robin Database and RRDTool
- latency and throughput, tradeoff between, Scalability Analysis
- least connections load balancing, Load Balancing
- least recently used (LRU) cache eviction algorithm, Cache efficiency: working sets and dynamic data
- Linux
- load balancers, Load Balancing-Applications of Monitoring
- load balancing
- load testing, Capacity Tracking Tools
- logging, aggregated configuration and management logging system, Goal 2: All Changes Happen in One Place
- logs
- long-term data storage, A real-world example: tracking storage consumption
- LRU (see least recently used cache eviction algorithm)
M
- Maglev load balancer, Load Balancing
- MAMP (Mac, Apache, MySQL, PHP) stack, Round-Robin Database and RRDTool
- master-slave database architecture, Finding database ceilings
- MEAN (MongoDB, Express.js, Angular.js, Node.js) stack, Round-Robin Database and RRDTool
- mean time to resolution (MTTR), User Expectations
- measurements, Make System Stats Tell Stories, Measurement: Units of Capacity-Database and Caching
- API usage and effect on capacity, API Usage and Its Effect on Capacity-API Usage and Its Effect on Capacity
- applications of monitoring, Applications of Monitoring-Special Use and Multiple Use Servers
- capacity tracking tools, Capacity Tracking Tools-Applications of Monitoring
- deciding which metrics to measure and follow, Measurement: Units of Capacity
- examples and reality, Examples and Reality
- of system resources, observer effect, Measurement: Units of Capacity
- tools for, Monitoring, Ad Hoc Measurement and Graphing Tools
- tools for, desired capabilities, Measurement: Units of Capacity
- usefulness in recognizing trends, Riding the Waves
- Memcached, Cache efficiency: working sets and dynamic data, The Effects of Increasing Capacity
- metrics
- microservice architecture, Predicting When Systems Will Fail
- microservices, Providing Measurement Points
- Microsoft Azure, Background
- mobile data traffic, growth of, Background
- mobile networks, speed of, External Service Monitoring
- monitoring, Application-Level Measurement
- MRTG tool, Network Measurement and Planning
- MSA (see microservice architecture)
- MySQL database, Predicting When Systems Will Fail, Database Capacity
P
- package-based installers, Automated Installation Tools
- paravirtual (PV), Automated Installation Tools
- peak-driven metrics, Trends, Curves, and Time
- peak-driven processing, Preliminaries
- percentage, scaling by, Scaling by Percentage-Illustration of Algorithm 2
- performance
- assessing website performance needs, Preliminaries
- designing fast and highly available, resources on, User Expectations
- determning requirements for, Setting Goals for Capacity
- maintaining, future needs for, Preliminaries
- monitoring by external services, External Service Monitoring
- perceived performance, metrics for, User Expectations
- performance tuning versus capacity planning, Performance and Capacity: Two Different Animals-Performance and Capacity: Two Different Animals
- requirements in SLAs, SLAs
- ping-pong effect, avoiding in autoscaling, Avoiding the ping-pong effect
- planning for capacity
- policies
- Postgres database, Database Capacity
- Pre-Boot Execution Environment (PXE), The Installation Process
- predicting trends (see trends, predicting)
- Preemptible VMs, The Challenge
- proactive scaling, Be proactive, not reactive
- problem identification, monitoring as tool for, Monitoring as a Tool for Urgent Problem Identification
- procurement, Buying Stuff-Buying Stuff, Procurement-Just-in-Time Inventory
- procurement pipeline, Procurement
- product and engineering teams, collaboration with, Make System Stats Tell Stories
- product planning, Application Usage Changes and Product Planning
- programmable infrastructure, Deployment
- public clouds, Background, Background
- PV (paravirtual) AMIs, Automated Installation Tools
R
- RAID, using on local disk subsystem, Storage I/O patterns
- read-only database operations, using replicated slave database, Predicting When Systems Will Fail
- Red Hat Kickstart, Automated Installation Tools, The Installation Process
- reliability, Make System Stats Tell Stories
- replaying requests, tools for, Finding web server ceilings in a load-balancing environment
- request rate limits, SLAs
- requests per second (RPS), Autoscaling on Amazon EC2, Startup Time Aware Scaling
- requirements, Different Kinds of Requirements and Measurements-User Expectations
- resources
- RESTful web APIs, Preliminaries
- revenue
- equating downtime to lost revenue, SLAs
- roles
- round-robin database (RRD), Fundamentals and Elements of Metric Collection Systems
- RPS (see requests per second)
- RRDTool, Round-Robin Database and RRDTool
- RUM, real user monitoring, Capacity Tracking Tools
S
- safety factors in web operations, Safety Factors
- SAS (Serial Attached SCSI) hard disk drives, Storage Capacity
- SATA protocol hard disks, Storage Capacity
- scalability analysis, Scalability Analysis
- scaling
- artificial (or load testing), Capacity Tracking Tools
- autoscaling (see autoscaling)
- database-driven web applications, Predicting When Systems Will Fail
- using public cloud infrastructure, Quick and Dirty Math
- vertical, horizontal, and diagonal scaling for hardware, Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)-Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)
- Scheduled Scaling (on AWS), Autoscaling on Amazon EC2
- schemas (database), optimization of, Database Capacity
- SCSI (Serial Computer System Interface) hard disk drives, Storage Capacity
- search, different intents of users, External Service Monitoring
- seasonal or holiday variations, Best Guesses
- server virtualization, Overview
- serverless architectures, Background
- servers
- comparing server architectures, Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)
- production load testing with a single machine, Finding web server ceilings in a load-balancing environment
- resource ceilings, Resource Ceilings
- serving more traffic with fewer servers, Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)
- special use and multiple use, capacity measurements in, Special Use and Multiple Use Servers, Special Use and Multiple Use Servers
- service level agreements (SLAs), SLAs-SLAs
- service-oriented architectures (SOAs), Background
- services
- shared-nothing architectures, horizontal scaling and, Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)
- Siege, A real-world example: cache measurement
- Simple Network Management Protocol (SNMP), Simple Network Management Protocol
- single point of failure, Hardware Decisions (Vertical, Horizontal, and Diagonal Scaling)
- single-server web application architecture, Providing Measurement Points
- SLAs (see service level agreements)
- smartphones, generations of, variations around the world, External Service Monitoring
- SNMP (Simple Network Management Protocol), Simple Network Management Protocol
- social media
- software, functionality moving from hardware to, Preliminaries
- software-defined networking (SDN), Preliminaries
- Solaris Jumpstart, Automated Installation Tools
- solid-state drives (SSDs), Storage Capacity
- Splunk, Treating Logs as Past Metrics
- SPOF (see single point of failure)
- Spot instances, market for in public clouds, The Challenge
- Squid, Cache efficiency: working sets and dynamic data
- startup time aware autoscaling, Startup Time Aware Scaling-Potpourri
- statistics, Make System Stats Tell Stories
- storage
- storage capacity, Storage Capacity-Database Capacity
- Storage-Area Network (SAN), Storage I/O patterns
- synthetic monitoring, Capacity Tracking Tools
- system statistics
- SystemImager, Automated Installation Tools, The Installation Process
..................Content has been hidden....................
You can't read the all page of ebook, please click
here login for view all page.