redis-server window版
资源内容介绍
redis-server window版 <link href="/image.php?url=https://csdnimg.cn/release/download_crawler_static/css/base.min.css" rel="stylesheet"/><link href="/image.php?url=https://csdnimg.cn/release/download_crawler_static/css/fancy.min.css" rel="stylesheet"/><link href="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89796021/raw.css" rel="stylesheet"/><div id="sidebar" style="display: none"><div id="outline"></div></div><div class="pf w0 h0" data-page-no="1" id="pf1"><div class="pc pc1 w0 h0"><img alt="" class="bi x0 y0 w1 h1" src="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89796021/bg1.jpg"/><div class="t m0 x1 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0">MSOpenTech’s Redis on Windows</div><div class="t m0 x1 h3 y2 ff2 fs1 fc1 sc0 ls0 ws0">We strive to have a stable, functionally equivalent and comparably performing version of Redis on </div><div class="t m0 x1 h3 y3 ff2 fs1 fc1 sc0 ls0 ws0">Windows. We have achieved performance nearly identical to the POSIX version running head-to-head </div><div class="t m0 x1 h3 y4 ff2 fs1 fc1 sc0 ls0 ws0">on identical hardware across the network. Aside from feature differences that help Redis take </div><div class="t m0 x1 h3 y5 ff2 fs1 fc1 sc0 ls0 ws0">advantage of the Windows infrastructure, our version of Redis should work in most situations with the </div><div class="t m0 x1 h3 y6 ff2 fs1 fc1 sc0 ls0 ws0">identical setup and configuration that one would use on a POSIX operating system.</div><div class="t m0 x1 h2 y7 ff1 fs0 fc0 sc0 ls0 ws0">How is Redis on Windows implemented?</div><div class="t m0 x1 h3 y8 ff2 fs1 fc1 sc0 ls0 ws0">Redis is a C code base that compiles under Visual Studio. Most of the code compiles with only minor </div><div class="t m0 x1 h3 y9 ff2 fs1 fc1 sc0 ls0 ws0">changes (due to syntactical differences between compilers and low-level API differences on Windows). </div><div class="t m0 x1 h3 ya ff2 fs1 fc1 sc0 ls0 ws0">There are a few areas where there are significant differences in how efficient Windows programs </div><div class="t m0 x1 h3 yb ff2 fs1 fc1 sc0 ls0 ws0">operate relative to POSIX programs. We have encapsulated most these differences in a platform specific </div><div class="t m0 x1 h3 yc ff2 fs1 fc1 sc0 ls0 ws0">library. The areas where there are significant differences are:</div><div class="t m0 x2 h3 yd ff2 fs1 fc1 sc0 ls0 ws0">•<span class="_ _0"> </span>Networking APIs </div><div class="t m0 x2 h3 ye ff2 fs1 fc1 sc0 ls0 ws0">•<span class="_ _0"> </span>POSIX File Descriptors</div><div class="t m0 x2 h3 yf ff2 fs1 fc1 sc0 ls0 ws0">•<span class="_ _0"> </span>POSIX fork()</div><div class="t m0 x2 h3 y10 ff2 fs1 fc1 sc0 ls0 ws0">•<span class="_ _0"> </span>Logging </div><div class="t m0 x2 h3 y11 ff2 fs1 fc1 sc0 ls0 ws0">•<span class="_ _0"> </span>Windows Services API</div><div class="t m0 x1 h4 y12 ff1 fs2 fc0 sc0 ls0 ws0">Networking Differences</div><div class="t m0 x1 h3 y13 ff2 fs1 fc1 sc0 ls0 ws0">The Windows networking stack is split between user mode code and kernel mode code. Transitions </div><div class="t m0 x1 h3 y14 ff2 fs1 fc1 sc0 ls0 ws0">between user and kernel mode are expensive operations. The POSIX networking APIs on Windows </div><div class="t m0 x1 h3 y15 ff2 fs1 fc1 sc0 ls0 ws0">utilize a programming model that incurs significant performance loss due to the kernel/user mode </div><div class="t m0 x1 h3 y16 ff2 fs1 fc1 sc0 ls0 ws0">transitions. Efficient Windows networking code instead uses the IO Completion Port model to reduce </div><div class="t m0 x1 h3 y17 ff2 fs1 fc1 sc0 ls0 ws0">the impact of this behavior. The APIs used and the programming model for IO Completion is different </div><div class="t m0 x1 h3 y18 ff2 fs1 fc1 sc0 ls0 ws0">enough that we were forced to implement a new networking layer in Redis.</div><div class="t m0 x1 h4 y19 ff1 fs2 fc0 sc0 ls0 ws0">File Descriptors</div><div class="t m0 x1 h3 y1a ff2 fs1 fc1 sc0 ls0 ws0">In a POSIX operating system, all data sources (files, pipes, sockets, mail slots, etc.) are referenced in code </div><div class="t m0 x1 h3 y1b ff2 fs1 fc1 sc0 ls0 ws0">with a handle called a file descriptor. These are low value integers that increment by one with each </div><div class="t m0 x1 h3 y1c ff2 fs1 fc1 sc0 ls0 ws0">successive file descriptor creation. All POSIX APIs that work with file descriptors will function without the </div><div class="t m0 x1 h3 y1d ff2 fs1 fc1 sc0 ls0 ws0">programmer having to know what kind of data source a file descriptor represents. On Windows, each </div><div class="t m0 x1 h3 y1e ff2 fs1 fc1 sc0 ls0 ws0">kind of data source has a separate kind of HANDLE. APIs that work with one HANDLE type will not work </div><div class="t m0 x1 h3 y1f ff2 fs1 fc1 sc0 ls0 ws0">with another kind of HANDLE. In order to make Redis operate with its assumptions about file descriptor </div><div class="t m0 x1 h3 y20 ff2 fs1 fc1 sc0 ls0 ws0">values and data source agnosticism, we implemented a Redis File Descriptor API layer.</div><div class="t m0 x1 h4 y21 ff1 fs2 fc0 sc0 ls0 ws0">fork()</div><div class="t m0 x1 h3 y22 ff2 fs1 fc1 sc0 ls0 ws0">The POSIX version of Redis uses the fork() API. There is no equivalent in Windows, and it is an </div><div class="t m0 x1 h3 y23 ff2 fs1 fc1 sc0 ls0 ws0">exceedingly difficult API to completely simulate. For most of the uses of fork() we have used Windows </div><div class="t m0 x1 h3 y24 ff2 fs1 fc1 sc0 ls0 ws0">specific programming idioms to bypass the need to use a fork()-like API. The one case where we could </div><div class="t m0 x1 h3 y25 ff2 fs1 fc1 sc0 ls0 ws0">not do so was with the point-in-time heap snapshot behavior that the Redis persistence model is based </div><div class="t m0 x1 h3 y26 ff2 fs1 fc1 sc0 ls0 ws0">on. We tried several different approaches to work around the need for a fork()-like API, but always ran </div><div class="t m0 x1 h3 y27 ff2 fs1 fc1 sc0 ls0 ws0">into significant performance penalties and stability issues. </div></div><div class="pi" data-data='{"ctm":[1.568627,0.000000,0.000000,1.568627,0.000000,0.000000]}'></div></div>