📄 vinum-data-integrity.html
字号:
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta name="generator" content="HTML Tidy, see www.w3.org" /><title>Data Integrity</title><meta name="GENERATOR" content="Modular DocBook HTML Stylesheet Version 1.7" /><link rel="HOME" title="FreeBSD Handbook" href="index.html" /><link rel="UP" title="The Vinum Volume Manager" href="vinum-vinum.html" /><link rel="PREVIOUS" title="Access Bottlenecks" href="vinum-access-bottlenecks.html" /><link rel="NEXT" title="Vinum Objects" href="vinum-objects.html" /><link rel="STYLESHEET" type="text/css" href="docbook.css" /></head><body class="SECT1" bgcolor="#FFFFFF" text="#000000" link="#0000FF" vlink="#840084"alink="#0000FF"><div class="NAVHEADER"><table summary="Header navigation table" width="100%" border="0" cellpadding="0"cellspacing="0"><tr><th colspan="3" align="center">FreeBSD Handbook</th></tr><tr><td width="10%" align="left" valign="bottom"><a href="vinum-access-bottlenecks.html"accesskey="P">Prev</a></td><td width="80%" align="center" valign="bottom">Chapter 17 The Vinum Volume Manager</td><td width="10%" align="right" valign="bottom"><a href="vinum-objects.html"accesskey="N">Next</a></td></tr></table><hr align="LEFT" width="100%" /></div><div class="SECT1"><h1 class="SECT1"><a id="VINUM-DATA-INTEGRITY" name="VINUM-DATA-INTEGRITY">17.4 DataIntegrity</a></h1><p>The final problem with current disks is that they are unreliable. Although disk drivereliability has increased tremendously over the last few years, they are still the mostlikely core component of a server to fail. When they do, the results can be catastrophic:replacing a failed disk drive and restoring data to it can take days.</p><p>The traditional way to approach this problem has been <span class="emphasis"><iclass="EMPHASIS">mirroring</i></span>, keeping two copies of the data on differentphysical hardware. Since the advent of the <acronym class="ACRONYM">RAID</acronym>levels, this technique has also been called <acronym class="ACRONYM">RAID level1</acronym> or <acronym class="ACRONYM">RAID-1</acronym>. Any write to the volume writesto both locations; a read can be satisfied from either, so if one drive fails, the datais still available on the other drive.</p><p>Mirroring has two problems:</p><ul><li><p>The price. It requires twice as much disk storage as a non-redundant solution.</p></li><li><p>The performance impact. Writes must be performed to both drives, so they take up twicethe bandwidth of a non-mirrored volume. Reads do not suffer from a performance penalty:it even looks as if they are faster.</p></li></ul><p>An alternative solution is <span class="emphasis"><iclass="EMPHASIS">parity</i></span>, implemented in the <acronymclass="ACRONYM">RAID</acronym> levels 2, 3, 4 and 5. Of these, <acronymclass="ACRONYM">RAID-5</acronym> is the most interesting. As implemented in Vinum, it isa variant on a striped organization which dedicates one block of each stripe to parity ofthe other blocks. As implemented by Vinum, a <acronym class="ACRONYM">RAID-5</acronym>plex is similar to a striped plex, except that it implements <acronymclass="ACRONYM">RAID-5</acronym> by including a parity block in each stripe. As requiredby <acronym class="ACRONYM">RAID-5</acronym>, the location of this parity block changesfrom one stripe to the next. The numbers in the data blocks indicate the relative blocknumbers.</p><p></p><div class="FIGURE"><a id="VINUM-RAID5-ORG" name="VINUM-RAID5-ORG"></a><p><b>Figure 17-3. RAID-5 Organization</b></p><p><img src="vinum/vinum-raid5-org.png" /></p></div><br /><br /><p>Compared to mirroring, <acronym class="ACRONYM">RAID-5</acronym> has the advantage ofrequiring significantly less storage space. Read access is similar to that of stripedorganizations, but write access is significantly slower, approximately 25% of the readperformance. If one drive fails, the array can continue to operate in degraded mode: aread from one of the remaining accessible drives continues normally, but a read from thefailed drive is recalculated from the corresponding block from all the remainingdrives.</p></div><div class="NAVFOOTER"><hr align="LEFT" width="100%" /><table summary="Footer navigation table" width="100%" border="0" cellpadding="0"cellspacing="0"><tr><td width="33%" align="left" valign="top"><a href="vinum-access-bottlenecks.html"accesskey="P">Prev</a></td><td width="34%" align="center" valign="top"><a href="index.html"accesskey="H">Home</a></td><td width="33%" align="right" valign="top"><a href="vinum-objects.html"accesskey="N">Next</a></td></tr><tr><td width="33%" align="left" valign="top">Access Bottlenecks</td><td width="34%" align="center" valign="top"><a href="vinum-vinum.html"accesskey="U">Up</a></td><td width="33%" align="right" valign="top">Vinum Objects</td></tr></table></div><p align="center"><small>This, and other documents, can be downloaded from <ahref="ftp://ftp.FreeBSD.org/pub/FreeBSD/doc/">ftp://ftp.FreeBSD.org/pub/FreeBSD/doc/</a>.</small></p><p align="center"><small>For questions about FreeBSD, read the <ahref="http://www.FreeBSD.org/docs.html">documentation</a> before contacting <<ahref="mailto:questions@FreeBSD.org">questions@FreeBSD.org</a>>.<br />For questions about this documentation, e-mail <<ahref="mailto:doc@FreeBSD.org">doc@FreeBSD.org</a>>.</small></p></body></html>⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -