1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
|
<html>
<head>
<title>diskstats moodss module</title>
</head>
<body>
<p>This is a view of the disks and/or partitions on a Linux machine, with statistics on read and write operations.
<br>On a 2.6 kernel, data is retrieved from the <i>/proc/diskstats</i> special file, whereas on a 2.4 kernel, it uses the <i>/proc/partitions</i> file, which requires the <b>per partition statistics in /proc/diskstats</b> feature to be enabled in the running kernel (that option is available with the 2.4.20 kernel version and above).
<p><i><b>Note</b>: see also the <b>disks</b> module for displaying more information on the disks themselves, and the <b>mounts</b> module to display the mounted file systems.</i>
<pre><img src="diskstats.gif"></pre>
<p>The initial table columns are:
<ul>
<li>disk or partition <b>name</b> (partitions end by a number)
<li><b>size</b> in megabytes (with first decimal below 100 megabytes, rounded above)
<li>number of kilobytes <b>read</b>, per second, averaged over the last poll period
<li>number of kilobytes <b>written</b>, per second, averaged over the last poll period
<li>time spent <b>reading</b> in percent during the last poll period (available for disks only)
<li>time spent <b>writing</b> in percent during the last poll period (available for disks only)
<li>how <b>busy</b> the disk is: time spent doing input/output in percent during the last poll period (available for disks only)
</ul>
<p>By default, only the disk entries are displayed, but that can be tuned using the options below.
<p><b>Module options:</b><ul>
<li><i>-a (--all)</i>
<br>Display both disk (such as <b>hda</b>, <b>sdb</b>, ...) and partition (such as <b>hda2</b>, <b>sdb1</b>, ...) entries (as shown in the screen shot above). Note: this option has a higher priority than the <i>--partitions</i> option.
<li><i>-C</i> <i>(only available for ssh on UNIX client)</i>
<br>Whether data compression is used on all data between client and server (useful on slow connections, see <i>ssh</i> manual for detailed information).
<li><i>-i file</i> <i>(only available for ssh on UNIX client)</i>
<br>Selects a file from which the identity (private key) for authentication is read (useful if an authentication agent is not running). It must not be protected by a passphrase.
<li><i>-p port</i> <i>(only available for ssh on UNIX client)</i>
<br>Port to connect to on the remote host.
<li><i>--partitions</i>
<br>Only display partition entries (with names such as <b>hda1</b>, <b>sdb1</b>, ...).
<li><i>-r (--remote) [[rsh|ssh]://][user@]host</i>
<br>Remote monitoring using <i>user</i> as logname on remote host <i>host</i> (<i>rsh</i> or <i>ssh</i> facilities must be properly setup). If <i>user</i> is not specified, current user is used as logname on remote host. The protocol is either <i>ssh</i> or <i>rsh</i> (used by default). The module title is set to <i>diskstats(host)</i>.
<br>When there is a communication error with the remote host, all rows disappear and the displayed table becomes empty. A descriptive error message is also generated in such a case.
</ul>
<p><b>Notes on remote monitoring:</b><ul>
<li>Using <i>ssh</i> is strongly recommended, as <i>rsh</i>, while being much less secure is also less efficient, as it requires creating a new session for each poll.
<li>On a Windows client, you must use the <i>putty</i> software package (see <i>install.txt</i>), which does not support <i>rsh</i> in non interactive sessions. Consequently, the <i>-r (--remote) ssh://session</i> syntax is required to remove any confusion (where <i>session</i> has been added to the running <i>pageant</i> application).
</ul>
<p><b>Examples:</b>
<pre>$ moodss diskstats --all
$ moodss diskstats --partitions
$ moodss diskstats -r jdoe@foo.bar.com
$ moodss diskstats -a -r ssh://jdoe@foo.bar.com
$ moodss diskstats --remote foo.bar.com</pre>
</body>
</html>
|
