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
|
TITLE=Installation DESCRIPTION=installing remstats
KEYWORDS=installation DOCTOP=index DOCPREV=required
DOCNEXT=install-user SECTION=Installation
How to install remstats
READ THE RELEASE NOTES FIRST. This page is generic and does NOT
include version-specific instructions.
I know that this is not simple. I do plan to make it simpler,
but it'll never be "`./configure; make; make install'" because I
don't know what you want to monitor. The two C programs (the
multiping manpage and the traceroute manpage) 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 `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.
1 Unpack the distribution tarball:
gunzip -dc remstats.tar.gz | tar xf -
2 create the remstats user and group, if you haven't already,
(by default `remstats' and `remstats' respectively.) (See also
the remstats user.)
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'
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 `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.
4 fix the config-base for site-specific things. Edit the following
files in /home/remstats/etc/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 /home/remstats/etc/config-base/* /home/remstats/etc/config-base/*/*
5 Make a config-dir 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 /home/remstats/etc
/home/remstats/bin/new-config config
/home/remstats/bin/new-ping-hosts groupname1 group1-hosts-file
/home/remstats/bin/new-ping-hosts groupname2 group2-hosts-file
...
/home/remstats/bin/new-port-hosts groupname3 port-hosts-file
/home/remstats/bin/new-snmp-hosts groupname4 SNMP-community-string snmp-hosts-file
After you've installed the the unix-status-server manpage on
some hosts, you can also use:
/home/remstats/bin/new-unix-hosts groupname5 unix-hosts-file
If you have any Windows NT hosts that you want to monitor,
after you have installed the the nt-status-server manpage,
you can run the nt-discover manpage 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.
6 Arrange for cron to run the run-remstats manpage 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 * * * * /home/remstats/bin/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 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.
7 [optional] Arrange for cron to run `do-traceroutes' at an appropriate
interval. You could run it in the wee hours of each morning like:
5 3 * * * /home/remstats/bin/do-traceroutes
This information isn't currently used, but I'm planning to
make use of it.
8 [optional] Arrange for cron to run the snmpif-description-updater manpage
periodically, if you have any snmpif-* RRDs, which you're likely to change
the descriptions on. Say every day, like:
0 3 * * * /home/remstats/bin/snmpif-description-updater
9 Arrange for cron to run the cleanup manpage every now and then to remove old
un-needed files, like:
0 3 * * * /home/remstats/bin/cleanup
This removes no-longer-needed files, like old host graphs,
traceroute results, log-files, ...
10 You'll need to set up your web-server to allow
CGI scripts in the remstats html tree and make sure that you're not
allowing everyone in.
11 Make a symlink in the html directory from whichever index
you prefer to index.cgi.
12 You'll want to look at the server installation docs
if you're going to be running any of the remote servers (
the log-server manpage, the nt-status-server manpage, the remoteping-server manpage, and
the unix-status-server manpage).
Enjoy your pretty pictures and I hope that you find them
usefull.
------------------------------------------------------------------
Last updated Mon Sep 10 15:18:26 EDT 2001 by <thomas.erskine@sourceworks.com>.
|
