构建高可用性系统的关键软件原则

2星需积分: 10 112 浏览量更新于2024-09-20 收藏 45KB PDF 举报

"设计高可用性系统软件原则(RTC_July_2002)" 在IT领域，构建高可用性（High-Availability, HA）系统是确保服务稳定性和连续性的关键任务。高可用性系统旨在容忍故障并能在不中断服务的情况下进行维护和升级。这一概念在电信行业中尤其重要，特别是在设计运营商级系统时，通常要求达到99.999%（即“五九”）的正常运行时间。高可用性的核心是故障管理，这包括如何限制故障的影响以及如何快速恢复。系统可用性的计算公式是：可用性 = (MTTF / (MTTF + MTTR)) 其中： MTTF（Mean Time To Failure）代表平均无故障时间， MTTR（Mean Time To Repair）代表平均修复时间。要提高系统的可用性，可以通过降低MTTR或增加MTTF来实现。在软件中，错误主要来源于MTTF，但通过分析方法计算这一数值通常是困难的。软件质量的提升通常伴随着更多的时间和金钱投入，以及更好的开发工具。然而，显著提高质量（即降低MTTF）往往成本高昂。相比之下，MTTR更可以通过改进流程、自动化和优化故障排查策略来有效计算和缩短。例如，采用冗余组件、故障隔离机制、自动故障切换和热备策略可以减少系统停机时间，从而降低MTTR。设计高可用性系统的一些关键原则包括： 1. 冗余与复制：通过部署冗余硬件和软件组件，确保在单个组件故障时，系统仍能继续运行。例如，使用负载均衡器分配流量，采用主-备或主-主数据库配置。 2. 故障隔离：设计系统时，应将不同功能模块相互隔离，使得一个模块的故障不会影响到整个系统。 3. 自动化恢复：实施监控和自动恢复机制，一旦检测到故障，立即执行预定的恢复流程，以快速修复问题。 4. 优雅降级：当系统部分功能无法正常工作时，允许其以降低功能但保持核心服务可用的方式运行。 5. 容错设计：设计能够容忍错误的系统架构，如使用容错算法和错误预测模型。 6. 持续监控：实时监控系统性能和状态，以便及时发现并解决问题。 7. 更新和维护：采用滚动更新或蓝绿部署等策略，使得系统升级可以在不影响用户的情况下进行。 8. 可测试性：设计易于测试的系统，确保所有组件在故障条件下都能按预期工作。 9. 文档与培训：提供详尽的操作手册和培训，确保运维人员在面对故障时能迅速响应。通过这些原则，开发者和运维人员可以构建出更加健壮、可靠的系统，满足高可用性的要求，确保业务的连续性和稳定性。

Reprinted from

July 2002

by Michael Christofferson, OSE Systems

igh availability (HA) requires building a system that is

tolerant to faults and that can be maintained and

enhanced without being taken out of service. From a

software design perspective, the key idea revolves around the

concept of fault management—how to contain faults and

recover from them. One of several classifications of “avail-

ability”, HA is the classification generally used in the telecom-

munications industry for carrier class systems. HA means that

the system features 99.999% uptime (“five nines”). The avail-

ability of a system is defined by the equation:

Where MTTF == Mean Time To Failure

And MTTR == Mean Time To Repair

The availability of a system may be improved by either

decreasing MTTR or increasing MTTF. MTTF in software,

from which most errors derive, is typically a difficult number

to calculate by any sort of analytic means. The quality of soft-

ware can almost always be improved by spending more time

and money, and by improving tools. Any significant improve-

ment in quality (MTTF), however, usually comes at a very

steep cost. MTTR on the other hand, can be computed much

more deterministically, and significant improvement can be

achieved at much lower relative and quantifiable cost. Most

HA recovery techniques focus on decreasing MTTR.

Errors, Faults and Failures

An error is an observable condition wherein a value or

response departs from the true or correct value. A fault is

defined as a failure of an interacting system or component that

causes an error. Afailure, then, is defined as a condition wherein

the observed behavior of the system or interacting component

deviates from the specified behavior. A fault causes an error,

which may lead to a failure. Or the error may be correctable

within the context of the detection of the error condition.

Further, a fault may cause another fault, which causes another

error and so on. But the important distinction is that errors are

not the fault.

Studies have shown, that try as we might, the exact source

of most faults cannot be completely located or determined.

Therefore, it often is not clear how extensive the “corruption”

that results from a fault may be. This means two things: First,

“repairing” the fault is often not possible. Second, recovery

from the error cannot be performed locally by the software that

detected the error, because the error handling software may

itself be corrupted. Some sort of more coherent system-wide

concept of fault management is required.

For most errors, we cannot effectively locate the exact

cause of the fault. Therefore, we try to contain the fault, local-

izing its effect, and thus preventing further damage by inducing

more faults and failures. Further, we need to determine the

exact collection of software or hardware units that must be

“repaired”. This requires a unified approach to fault handling or

Effective design for high availability requires a coherent, system-wide

concept or model of fault management. Some requirements for

software design are derived from such a model.

Software Principles for Designing

High-Availability Systems

Fault Management

Software&DevelopmentTools

MTTF

(MTTF + MTTR)

Availability =

下载后可阅读完整内容，剩余3页未读，立即下载

哦布斯

粉丝: 13
资源: 1

构建高可用性系统的关键软件原则

Pulsed_Power_Systems__Principles_and_Applications__Power_Systems_.pdf

计算机图形学基础教程电子书链接

computational fluid dynamics - principles and applications - j. blazek.pdf

写20个研发云项目设计软件的性能要求，对电脑配置的，带参数。比如打开设计软件小于10秒，系统宕机恢复时间小于半小时

principles of adaptive filters and self-learning systems

双光子荧光显微镜相关文献

介绍机器人路径规划的外文文献

principles of planar near-field antenna measurements

最新资源