|
Lines 55-61
Link Here
|
| 55 |
|
55 |
|
| 56 |
|
56 |
|
| 57 |
<para>Disks are getting bigger, but so are data storage requirements. |
57 |
<para>Disks are getting bigger, but so are data storage requirements. |
| 58 |
Often you ill find you want a file system that is bigger than the disks |
58 |
Often you will find you want a file system that is bigger than the disks |
| 59 |
you have available. Admittedly, this problem is not as acute as it was |
59 |
you have available. Admittedly, this problem is not as acute as it was |
| 60 |
ten years ago, but it still exists. Some systems have solved this by |
60 |
ten years ago, but it still exists. Some systems have solved this by |
| 61 |
creating an abstract device which stores its data on a number of disks.</para> |
61 |
creating an abstract device which stores its data on a number of disks.</para> |
|
Lines 70-76
Link Here
|
| 70 |
disks.</para> |
70 |
disks.</para> |
| 71 |
|
71 |
|
| 72 |
<para>Current disk drives can transfer data sequentially at up to |
72 |
<para>Current disk drives can transfer data sequentially at up to |
| 73 |
30 MB/s, but this value is of little importance in an environment |
73 |
70 MB/s, but this value is of little importance in an environment |
| 74 |
where many independent processes access a drive, where they may |
74 |
where many independent processes access a drive, where they may |
| 75 |
achieve only a fraction of these values. In such cases it is more |
75 |
achieve only a fraction of these values. In such cases it is more |
| 76 |
interesting to view the problem from the viewpoint of the disk |
76 |
interesting to view the problem from the viewpoint of the disk |
|
Lines 85-94
Link Here
|
| 85 |
|
85 |
|
| 86 |
<para><anchor id="vinum-latency"> |
86 |
<para><anchor id="vinum-latency"> |
| 87 |
Consider a typical transfer of about 10 kB: the current generation of |
87 |
Consider a typical transfer of about 10 kB: the current generation of |
| 88 |
high-performance disks can position the heads in an average of 6 ms. The |
88 |
high-performance disks can position the heads in an average of 3.5 ms. The |
| 89 |
fastest drives spin at 10,000 rpm, so the average rotational latency |
89 |
fastest drives spin at 15,000 rpm, so the average rotational latency |
| 90 |
(half a revolution) is 3 ms. At 30 MB/s, the transfer itself takes about |
90 |
(half a revolution) is 1.75 ms. At 30 MB/s, the transfer itself takes about |
| 91 |
350 μs, almost nothing compared to the positioning time. In such a |
91 |
150 μs, almost nothing compared to the positioning time. In such a |
| 92 |
case, the effective transfer rate drops to a little over 1 MB/s and is |
92 |
case, the effective transfer rate drops to a little over 1 MB/s and is |
| 93 |
clearly highly dependent on the transfer size.</para> |
93 |
clearly highly dependent on the transfer size.</para> |
| 94 |
|
94 |
|
|
Lines 151-157
Link Here
|
| 151 |
For example, the first 256 sectors may be stored on the first disk, the |
151 |
For example, the first 256 sectors may be stored on the first disk, the |
| 152 |
next 256 sectors on the next disk and so on. After filling the last |
152 |
next 256 sectors on the next disk and so on. After filling the last |
| 153 |
disk, the process repeats until the disks are full. This mapping is called |
153 |
disk, the process repeats until the disks are full. This mapping is called |
| 154 |
<emphasis>striping</emphasis> or RAID-0. |
154 |
<emphasis>striping</emphasis> or <acronym>RAID-0</acronym>. |
| 155 |
|
155 |
|
| 156 |
<footnote> |
156 |
<footnote> |
| 157 |
<indexterm> |
157 |
<indexterm> |
|
Lines 250-256
Link Here
|
| 250 |
</figure> |
250 |
</figure> |
| 251 |
</para> |
251 |
</para> |
| 252 |
|
252 |
|
| 253 |
<para>Compared to mirroring, RAID-5 has the advantage of requiring |
253 |
<para>Compared to mirroring, <acronym>RAID-5</acronym> has the advantage of requiring |
| 254 |
significantly less storage space. Read access is similar to that of |
254 |
significantly less storage space. Read access is similar to that of |
| 255 |
striped organizations, but write access is significantly slower, |
255 |
striped organizations, but write access is significantly slower, |
| 256 |
approximately 25% of the read performance. If one drive fails, the array |
256 |
approximately 25% of the read performance. If one drive fails, the array |
|
Lines 470-476
Link Here
|
| 470 |
the system automatically assigns names derived from the plex name by |
470 |
the system automatically assigns names derived from the plex name by |
| 471 |
adding the suffix <emphasis>.s</emphasis><emphasis>x</emphasis>, where |
471 |
adding the suffix <emphasis>.s</emphasis><emphasis>x</emphasis>, where |
| 472 |
<emphasis>x</emphasis> is the number of the subdisk in the plex. Thus |
472 |
<emphasis>x</emphasis> is the number of the subdisk in the plex. Thus |
| 473 |
Vinum gives this subdisk the name <emphasis>myvol.p0.s0</emphasis></para> |
473 |
Vinum gives this subdisk the name <emphasis>myvol.p0.s0</emphasis>.</para> |
| 474 |
</listitem> |
474 |
</listitem> |
| 475 |
</itemizedlist> |
475 |
</itemizedlist> |
| 476 |
|
476 |
|
|
Lines 736-743
Link Here
|
| 736 |
</listitem> |
736 |
</listitem> |
| 737 |
|
737 |
|
| 738 |
<listitem> |
738 |
<listitem> |
| 739 |
<para>The directories <devicename>/dev/vinum/plex</devicename> and |
739 |
<para>The directories <devicename>/dev/vinum/plex</devicename>, |
| 740 |
<devicename>/dev/vinum/sd</devicename>, |
740 |
<devicename>/dev/vinum/sd</devicename>, and |
| 741 |
<devicename>/dev/vinum/rsd</devicename>, which contain block device |
741 |
<devicename>/dev/vinum/rsd</devicename>, which contain block device |
| 742 |
nodes for each plex and block and character device nodes respectively |
742 |
nodes for each plex and block and character device nodes respectively |
| 743 |
for each subdisk.</para> |
743 |
for each subdisk.</para> |