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
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
|
=cut
TITLE=Installation
DESCRIPTION=installing remstats
KEYWORDS=installation
DOCTOP=index
DOCPREV=required
DOCNEXT=install-user
SECTION=Installation
=pod
=head1 How to install remstats
READ THE L<RELEASE NOTES|releasenotes.html> FIRST. This page is generic and does
B<NOT> include version-specific instructions.
I know that this is not simple. I do plan to make it simpler, but it'll B<never>
be "C<./configure; make; make install>" because I don't know what you want
to monitor.
The two C programs (L<multiping> and L<traceroute>) now use autoconf, and the
main configure script works (from the outside) simlarly to an autoconf-generated
configure. I haven't seen a need to convert it to autoconf yet.
It's mostly perl scripts and
if you have the right version of perl properly installed, it shouldn't need
anything special. The C<unix-status-server> is a slight exception to this,
but the only configuration needed so far is done dynamically and is only the
location of the various required utilities.
=over 4
=item 1 Unpack the distribution tarball:
gunzip -dc remstats.tar.gz | tar xf -
=item 2 create the remstats user and group, if you haven't already,
(by default C<remstats> and C<remstats> respectively.) (See also
L<the remstats user|install-user>.)
=for html <P></P>
=item 3 Build and install the software. If you're upgrading, you
might want to take a copy of fixup.config from the old version:
sh configure
If you want to override the defaults, then run
sh configure --help
for a list of what can be overridden.
[Check fixup.config to make sure it is properly setup.]
make all
make install
su -c 'make install-suid'
B<Note:> this step also customizes the programs and documentation
with your choice of directories, owner, ... so this documentation
should refer to your setup after you've done the install.
The C<make install-suid> simply makes traceroute and multiping suid root.
They won't work most places unles run as root, one way or another. Since I
don't like to run all of remstats as root, this was the best compromise I
could come up with.
=item 4 fix the config-base for site-specific things. Edit the following
files in @@ETCDIR@@/config-base, looking the the string "FIXME", without the "quotes".
alerts general html scripts/http-proxy
I'll try to keep this list up to date, but you can make sure by doing:
grep -l FIXME @@ETCDIR@@/config-base/* @@ETCDIR@@/config-base/*/*
=item 5 Make a L<config-dir|configuration> to describe what you
want to monitor. You can do this by hand, or using the configuration
building tools. To use the tools, you'll have to make a few files
listing various kinds of hosts:
cd @@ETCDIR@@
@@BINDIR@@/new-config config
@@BINDIR@@/new-ping-hosts groupname1 group1-hosts-file
@@BINDIR@@/new-ping-hosts groupname2 group2-hosts-file
...
@@BINDIR@@/new-port-hosts groupname3 port-hosts-file
@@BINDIR@@/new-snmp-hosts groupname4 SNMP-community-string snmp-hosts-file
After you've installed the L<unix-status-server> on some hosts, you can also
use:
@@BINDIR@@/new-unix-hosts groupname5 unix-hosts-file
If you have any Windows NT hosts that you want to monitor, after you
have installed the L<nt-status-server>, you can run L<nt-discover> to
find and add the NT hosts for a given NT domain.
If you're going to use the log-collector, you'll have to build the
rrd entries for each by hand. There doesn't seem to be much standard
in where log-files go, let alone what's in them.
=item 6 Arrange for cron to run L<run-remstats> at an appropriate interval.
For a five-minute interval, something like the following will do:
0,5,10,15,20,25,30,35,40,45,50,55 * * * * @@BINDIR@@/run-remstats
This checks the configuration, collects the new data, updates the rrds,
runs the monitors to compute statuses and updates the web-pages. Note: it
does B<not> re-write the web-pages for every iteration; it only does so when
configuration files have changed, as the web-pages will show new data by
themselves.
=item 7 [optional] Arrange for cron to run C<do-traceroutes> at an appropriate
interval. You could run it in the wee hours of each morning like:
5 3 * * * @@BINDIR@@/do-traceroutes
This information isn't currently used, but I'm planning to make use of it.
=item 8 [optional] Arrange for cron to run L<snmpif-description-updater>
periodically, if you have any snmpif-* RRDs, which you're likely to change
the descriptions on. Say every day, like:
0 3 * * * @@BINDIR@@/snmpif-description-updater
=item 9 Arrange for cron to run L<cleanup> every now and then to remove old
un-needed files, like:
0 3 * * * @@BINDIR@@/cleanup
This removes no-longer-needed files, like old host graphs, traceroute results,
log-files, ...
=item 10 You'll need to set up your L<web-server|install-webserver> to allow
CGI scripts in the remstats html tree and make sure that you're not
allowing everyone in.
=for html <P></P>
=item 11 Make a symlink in the html directory from whichever index
you prefer to index.cgi.
=for html <P></P>
=item 12 You'll want to look at the L<server installation docs|install-servers>
if you're going to be running any of the remote servers (
L<log-server>, L<nt-status-server>, L<remoteping-server>, and
L<unix-status-server>).
=back 4
Enjoy your pretty pictures and I hope that you find them usefull.
|
