Attention: This document covers installations up to and including NMIS Version 8.5.4G and will not be updated in the future.
For NMIS versions 8.5.6G and newer the installation process was simplified substantially, and there is an updated version of the installation guide for 8.5.6G.
| Table of Contents |
|---|
Introduction
This document will describe how to install NMIS8 on GNU/Linux distros. This document is based on Terrell Prude Jr’s document for the installation of NMIS 4.3.x, v1.1 April 1, 2010.
Caveats
NMIS8 has been widely deployed and tested on many Linux distros, with Red Hat/CentOS being the most popular. This document contains variations specific to several Linux distros, but not all of them have been tested equally heavily. Please provide feedback and submit modifications or changes to support@opmantek.com.
Assumptions
The reader is assumed to have a working knowledge of UNIX or GNU/Linux systems, how to compile software, and what Perl is. While NMIS is written in Perl, you do not need to be a "Perl hacker" to install this program and make it work.
Also, this document will use the <nmisdir> tag to refer to NMIS's installation directory. The default location is /usr/local/nmis8.
Installation
Linux Distribution
First, install your GNU/Linux distribution. With both CentOS 5 and Debian, I do a basic installation with no extra stuff. That means no GNOME, KDE, "Server" packages, or anything else. Just a base installation is all that's necessary. This guide has been checked on a CentOS minimal install which is indeed quite minimal, because of this you may see some packages being installed that you might normally assume to be installed.
Linux Security Settings
If you're running CentOS or any other Red Hat-derived distribution, be sure that SELinux is turned off, Permissive mode was tested and it worked well, just very nagging. So disabling SELinux is a lot easier that way.
Also, I like to turn off the default Red Hat firewall. It is assumed that the reader knows how to do both of these; if not, there are plenty of resources on the Web describing how.
Additional Packages
Now that you have your system set up, there are some things that we need to install to make NMIS work. Some of these are required for the compilation of certain Perl modules and of RRDtool. These packages will come from your distribution's repositories.
CentOS Additional Packages:
| Code Block |
|---|
su - root # or sudo sh yum install autoconf automake gcc cvs yum install cairo cairo-devel \ pango pango-devel \ glib glib-devel \ libxml2 libxml2-devel \ gd gd-devel libXpm-devel libXpm yum install openssl openssl-devel yum install net-snmp net-snmp-lib net-snmp-utils net-snmp-perl perl-Net-SSLeay yum install httpd fping yum install make groff cpan crontabs yum -y install dejavu* perl-Authen-SASL |
Debian/Ubuntu Additional Packages:
| Code Block |
|---|
su root # or sudo sh apt-get install autoconf automake gcc make \ libcairo2 libcairo2-dev libglib2.0-dev libpango1.0-dev libxml2 libxml2-dev libgd-gd2-perl \ libnet-ssleay-perl libcrypt-ssleay-perl apache2 fping \ snmp snmpd libnet-snmp-perl libcrypt-passwdmd5-perl \ libjson-xs-perl libnet-dns-perl libio-socket-ssl-perl libwww-perl libnet-smtp-ssl-perl \ libcrypt-unixcrypt-perl libsoap-lite-perl libdata-uuid-perl \ libproc-processtable-perl libnet-ldap-perl libnet-snpp-perl libdbi-perl libtime-modules-perl libauthen-sasl-perl |
RRDTool Installation
We now need to get RRDtool, by Tobi Oetiker. The NMIS development team suggests that you download the latest stable RRDtool and compile it yourself - however, for normal NMIS deployments the distribution-provided RRDtool is generally sufficient.
If you want to skip the build step and use packaged versions, simply install the following packages and then skip to the next step.
Debian/Ubuntu:
| Code Block |
|---|
apt-get install rrdtool librrds-perl |
CentOS/Red Hat:
There are suitably recent RRDTool packages for CentOS, but they are not part of the 'default' repositories (these don't contain anything or very outdated versions). To work around this you need to enable the RepoForge repository first to gain access.
| Code Block |
|---|
# adding repoforge for centos 6: yum install http://pkgs.repoforge.org/rpmforge-release/rpmforge-release-0.5.3-1.el6.rf.x86_64.rpm # centos 6 ships an ancient rrdtool, repoforge-extras has what we're after yum -y --enablerepo=rpmforge-extras install rrdtool rrdtool-perl |
Compiling your own RRDtool:
The latest version of RRD is available at http://oss.oetiker.ch/rrdtool/.
Unpack the RRDtool tarball. NMIS8 has been testing using v1.4.7 and for this example. To build and install a “GNU” make based system, you run ./configure, then make, then make install, for RRDTool for NMIS we need a configure option setup.
| Code Block |
|---|
tar -xvf rrdtool-1.4.8.tar.gz [local]# cd rrdtool-1.4.8 [rrdtool-1.4.8]# ./configure --enable-shared --prefix=/usr/local/rrdtool |
When this step is finished, you should see something like this:
| Code Block |
|---|
ordering CD from http://tobi.oetiker.ch/wish .... just kidding ;-) ---------------------------------------------------------------- Config is DONE! With MMAP IO: yes Build rrd_getopt: no Static programs: no Perl Modules: perl_piped perl_shared Perl Binary: /usr/bin/perl Perl Version: 5.8.8 Perl Options: PREFIX=/usr/local/rrdtool LIB=/usr/local/rrdtool/lib/perl/5.8.8 Ruby Modules: Ruby Binary: no Ruby Options: sitedir=/usr/local/rrdtool/lib/ruby Build Lua Bindings: no Build Tcl Bindings: no Build Python Bindings: no Build rrdcgi: yes Build librrd MT: yes Use gettext: yes With libDBI: no Libraries: -lxml2 -lcairo -lcairo -lcairo -lm -lcairo -lpng12 -lglib-2.0 -lpangocairo-1.0 -lpango-1.0 -lcairo -lgobject-2.0 -lgmodule-2.0 -ldl -lglib-2.0 Type 'make' to compile the software and use 'make install' to install everything to: /usr/local/rrdtool. ... that wishlist is NO JOKE. If you find RRDtool useful make me happy. Go to http://tobi.oetiker.ch/wish and place an order. -- Tobi Oetiker <tobi@oetiker.ch> ---------------------------------------------------------------- |
So far, so good. You have all the things RRDtool needs to compile. Let's keep going.
| Code Block |
|---|
make && make install |
Now, rrdtool should be installed the way NMIS likes it.
NMIS8 Installation
And speaking of NMIS, if you haven't already downloaded it, the latest version (v8.5G at time of this writing) is available at http://opmantek.com
| Code Block |
|---|
cd ~ tar -xvzf nmis8.5g.tar.gz |
nmis8.5g, which can be installed manually or semi-automatically.Automatic Installation
install.pl.
| Code Block |
|---|
cd nmis8.5g ./install.pl -h ...shows the list of options for the install script ./install.pl listdeps=true ...shows what perl modules are missing, and how to get them /install.pl ...simply follow the on-screen prompts to perform the actual installation. |
The installer will likely tell you that a number of required CPAN modules are missing and have to be installed, but the on-screen prompts will guide you through that process.
Manual Installation
At this point, I like to copy this entire directory to the /usr/local directory tree and make some symlinks, like so.
| Code Block |
|---|
mkdir /usr/local/nmis8/ cp -r nmis8.4.8g/* /usr/local/nmis8/ rm -rf ~/nmis8.4.8g cd /usr/local |
My <nmisdir> is, in this case, /usr/local/nmis8. For those who have been using NMIS for a while, make a /usr/local/nmis.
| Code Block |
|---|
ln -s nmis8 nmis |
The <nmisdir> tree needs to be owned by the user "nmis", which doesn't exist yet. Furthermore, the userID under which Apache runs ("apache" on Red Hat and "www-data" on Debian) needs to have access to the <nmisdir> tree. So, we need to add to make these things happen. We are still in the /usr/local directory at this point.
| Code Block |
|---|
useradd nmis groupadd nmis #this may return "groupadd: group 'nmis' already exists" this can safely be ignored but if you want to verify try this: groups nmis |
Now we add the Apache userID to the nmis group.
| Code Block |
|---|
usermod -G nmis apache # Red Hat/CentOS adduser nmis www-data # Debian/Ubuntu |
NOTE: the stuff that used to be sitting in <nmisdir>/conf in previous NMIS tarballs is now located in <nmisdir>/install. Since we're doing a new installation here, we'll just copy everything in <nmisdir>/install to <nmisdir>/conf.
| Code Block |
|---|
cd /usr/local/nmis/ cp -rPv install/* conf/ # also copy models-install to models cp models-install/* models/ |
Perl CPAN Module Installation
Now, let's get busy with the Perl modules. NMIS, being written in Perl, uses numerous "Perl modules". These are available from the Comprehensive Perl Archive Network, known affectionately as "CPAN". No, not "C-SPAN", I mean "CPAN". :-)
Here's how we get to CPAN.
| Code Block |
|---|
perl -MCPAN -e "shell" # alternatively, with more modern cpan installations: cpan |
You will be asked a bunch of questions. It is safe to select all the defaults. Indeed, if you don't know what the questions are asking you, selecting the defaults is strongly recommended! HOWEVER, with some distros (Debian is a notable exception) you must select some CPAN mirrors, because that's where you're going to download your Perl modules from. Choose those which are closest to you. For me, that's North America; for Peruvians and Brazilians, that's South America; and so on.
On some distros like Red Hat/CentOS, you should start with this command. Debian doesn't seem to require it, which is good, because the command doesn't work on Debian's CPAN client anyway. :-)
| Code Block |
|---|
$cpan> o conf set LC_ALL=C #(Red Hat/CentOS) |
Additionally, I like to ensure my CPAN client program, which is what we're running ("CPAN.pm"), is the latest, up-to-date version. This will take a while and ask for several other "dependency" modules. Just tell it "yes" each time. If you would like to have cpan automatically choose yes use this code:
| Code Block |
|---|
$cpan> o conf prerequisites_policy follow $cpan> o conf commit #use this if you want to save this setting for next time |
And run this code to do the update:
| Code Block |
|---|
$cpan> $cpan> reload cpan |
OK, we've got CPAN prepped, locked, and loaded. :-) Now, install all the Perl modules listed in the official NMIS docs, like so.
$cpan> (assume the following lines are at this prompt)
| Code Block |
|---|
install JSON install JSON::XS install Proc::ProcessTable install Image::Resize install Net::SNMP install Time::HiRes install Time::ParseDate install Statistics::Lite install Net::Syslog install Net::SNPP install Net::LDAP install Net::SMTP install Net::SMTP::SSL install Net::DNS # Only required if using DNS lookups for geoplots install IPC::Shareable # Only required if multithreading on Unix/Linux. install BSD::Resource # Only required if multithreading on Unix/Linux. install GD::Graph # Only required for Capacity Planning tool. install Net::EasyTCP # Only required for master slavepoller operation. install Data::Dumper # Required for cgi metrics stats caching install Cache::Mmap # required for caching cgi metric stat install Proc::Queue install Date::Parse install Date::Calc install Math::Round install DBI install CGI::Pretty install Crypt::DES #Now, you're ready to exit CPAN. exit |
With recent versions of cpan it's also possible to give the modules to install on the command line:
| Code Block |
|---|
cpan Math::Round DBI CGI::Pretty # ...and so on |
Unix File Permissions
Make nmis user and group own all the files.
| Code Block |
|---|
chown -Rv nmis:nmis <nmisdir> chmod -Rv g+w <nmisdir> |
If you decide on a different user and group name, update the NMIS Configuration Config.nmis accordingly.
One of the joys of Unix is granular file permissions, one of the frustrations of Unix is granular file permissions. To assist we have added a handy script in the admin directory <nmisdir>/admin/fixperms.pl which will read the NMIS Configuration and fix the permissions accordingly.
| Code Block |
|---|
admin/fixperms.pl # This script will fix the permissions for NMIS based on the configuration /usr/local/nmis8/admin/../conf/Config.nmis # The directory to be processed is: /usr/local/nmis8 # The user will be set to: nmis # The group will be set to: nmis |
Initial NMIS Test
You should now be set. As an initial check, I like to go into my <nmisdir>/cgi-bin directory and run
| Code Block |
|---|
cd /usr/local/nmis8/cgi-bin ./nmiscgi.pl |
and see what happens. If all is working, you should get a bunch of HTML code. If you don't, it'll probably be NMISCGI complaining that it cannot find a given Perl module. Review your steps above; you likely missed something. The important thing to do is not to panic.
Initial NMIS Troubleshooting
RRDTool Libraries not Found
(This should no longer be a problem on NMIS8, but if you did something different with RRDTool, this is an easy way to fix it.)
For some reason, on some distros, NMIS will not be able to find the file RRDs.pm, which is a Perl module that comes with RRDtool, and that NMIS needs to be able to find and use. I ran into this problem. Fortunately, the NMIS development team included a small hack to take care of this. The line explicitly "uses" the RRDs.pm file in the/usr/local/rrdtool directory tree, but you have to uncomment that line in the NMIS program files themselves.
One way to quickly accomplish that is to run the following script within /usr/local/nmis8/bin and /usr/local/nmis8/cgi-bin. You can just cut 'n' paste it into your terminal.
| Code Block |
|---|
#!/bin/bash for file in *.pl do sed -i 's|#use lib "/usr/local/rrdtool|use lib "/usr/local/rrdtool|g' $file done |
NMIS Config Test
Once this is working, you're ready to head into the <nmisdir>/bin directory and do the "acid" test.
| Code Block |
|---|
cd /usr/local/nmis8 ./bin/nmis.pl type=config debug=true |
You should get what looks like a long checklist with "OK" at the end of every check. If you do this, then NMIS itself is configured correctly. The output will look something like this:
| Code Block |
|---|
./bin/nmis.pl type=config debug=true NMIS version 8.1.1 Root directory of NMIS is /usr/local/nmis8 05:20:05 checkConfig, Config Checking - Checking database directories 05:20:05 loadTable, ERROR file does not exist dir=var name=nmis-event 05:20:05 writeHashtoFile, write data to /usr/local/nmis8/var/nmis-event.nmis 05:20:05 loadTable, ERROR file does not exist dir=var name=nmis-system 05:20:05 writeHashtoFile, write data to /usr/local/nmis8/var/nmis-system.nmis Continue with bin/nmis.pl type=apache for configuration rules of the Apache web server |
The errors above are OK, NMIS will fix them itself.
NMIS Web Server Setup
Now, we need to set up the Web server so that folks can actually *get to* NMIS and look at the data. :-) That means generating an Apache config file pointing to the NMIS directory tree. NMIS itself is smart enough to be able to generate this for us. Thanks, NMIS team!
All you have to do is this.
| Code Block |
|---|
./bin/nmis.pl type=apache > 00nmis.conf |
We name the file prepending 00 so that it is the first config file read. The main part of the Apache config is as follows.
| Code Block |
|---|
# NMIS Aliases Alias /nmis8/ "/usr/local/nmis8/htdocs/" <Directory ""> Options Indexes FollowSymLinks MultiViews AllowOverride None Order allow,deny Allow from all </Directory> Alias /menu8/ "/usr/local/nmis8/menu/" <Directory "/menu8"> Options Indexes FollowSymLinks MultiViews AllowOverride None Order allow,deny Allow from all </Directory> ScriptAlias /cgi-nmis8/ "/usr/local/nmis8/cgi-bin/" <Directory "/usr/local/nmis8/cgi-bin"> Options +ExecCGI Order allow,deny Allow from all </Directory> |
You can use whatever filename you want instead of "00nmis.conf". This is very distro-dependent.
CentOS and other Red Hat derivatives:
Copy 00nmis.conf to /etc/httpd/conf.d/
| Code Block |
|---|
cp ./00nmis.conf /etc/httpd/conf.d/ |
Debian/Ubuntu:
Copy 00nmis.conf /etc/apache2/sites-available/, and enable that site:
| Code Block |
|---|
cp ./00nmis.conf /etc/apache/sites-available/nmis a2ensite nmis |
You could also integrate the 00nmis.conf into your existing web server configuration under its own virtual host, or into an existing set of hosts. This means you can run other version of NMIS on the same server, or for different customers for example, this is intended to be flexible for alternate integration.
Now that your NMIS config file for Apache is present--and if necessary, tweaked for your distro--start Apache itself. If it's already started, then restart it. How to do this is also very distro-dependent. Unlike me, BSD and Solaris users don't need no steenkin' startup scripts. :-)
Auto Start HTTPD Server
Ensure that you tell Linux to start HTTPD automatically
| Code Block |
|---|
/sbin/chkconfig httpd on |
Debian/Ubuntu Users
Nothing should be necessary as the apache2 is set to automatically start during the initial package installation, but if this should have failed:
| Code Block |
|---|
update-rc.d apache2 defaults |
Start HTTPD Server
You will need to start or restart Apache after configuration.
CentOS/Red Hat:
| Code Block |
|---|
/etc/init.d/httpd start # or /etc/init.d/httpd restart |
Debian/Ubuntu:
| Code Block |
|---|
service apache2 start # or service apache2 restart |
If you get an error like this, it just means the hostname and FQDN on the Linux box hasn’t been setup right, verify you have this setup correctly and update the httpd.conf for Apache accordingly.
| Code Block |
|---|
Starting httpd: httpd: apr_sockaddr_info_get() failed for nmis-dev httpd(apache2) : Could not reliably determine the server's fully qualified domain name, using 127.0.0.1 for ServerName |
Ensure Security settings will allow you to access the web server
For the purpose of this guide we will just disable security settings, this may not be the correct answer for you depending on your situation
IPTables
We will just disable ip tables, on Redhat/CentOS
| Code Block |
|---|
service iptables stop #make sure the service isn't running chkconfig iptables off |
SELinux
Again, just turn it off. Edit /etc/selinux/config
| Code Block |
|---|
#change this line: SELINUX=enforcing #to this: SELINUX=disabled |
You can reboot now, or you can step SELINUX with the following command.
| Code Block | ||
|---|---|---|
| ||
echo 0 >/selinux/enforce |
Quick Testing
Now, fire up your favourite Web browser (mine's Konqueror) and surf to your server. The URL will be "http://yourhostname.domain.tld/cgi-nmis8/nmiscgi.pl". Of course, if you have no DNS or hosts lookup for this box yet, just replace "yourhostname.domain.tld" with your new NMIS server's IP address.
You should get the NMIS Dashboard. If you don't, then check to make sure you included the "/cgi-nmis8/nmiscgi.pl" in the URL. Yep, I've made that mistake too many times to count. It's easy to forget, so make sure you include it.
Congratulations! It works, and you're almost done. We're in the home stretch.
NMIS Crontab Configuration
We ensured crontab was installed above, now we'll make sure it's running:
| Code Block |
|---|
chkconfig crond on /etc/init.d/crond start |
Debian/Ubuntu: nothing needs to be done. Cron is automatically started.
Now we have to tell NMIS to run the data collection every so often. For that, we use a cron job. Edit root's crontab ("crontab -e" at the root shell), like so. Remember to change the first line, the "MAILTO" line, to contain your actual email address.
| Code Block |
|---|
MAILTO=WhoeverYouAre@yourdomain.tld ###################################################### # NMIS8 Cronfig ###################################################### # Run Statistics Collection */5 * * * * /usr/local/nmis8/bin/nmis.pl type=collect mthread=true maxthreads=10 ###################################################### # Run Summary Update more frequently */3 * * * * /usr/local/nmis8/bin/nmis.pl type=summary ##################################################### # Run the interfaces 4 times an hour with Thresholding on!!! */15 * * * * nice /usr/local/nmis8/bin/nmis.pl type=threshold ###################################################### # Run the update once a day 30 20 * * * nice /usr/local/nmis8/bin/nmis.pl type=update mthread=true maxthreads=10 ###################################################### # Check to rotate the logs 4am every day UTC 5 20 * * * /usr/sbin/logrotate /usr/local/nmis8/conf/logrotate.conf ################################################## # save this crontab every day 0 8 * * * crontab -l > /usr/local/nmis8/conf/crontab.root # These file locations for are Red Hat. #0 8 * * * cp /etc/httpd/conf/httpd.conf /usr/local/nmis8/conf/httpd.conf.backup #0 8 * * * cp /etc/httpd/conf.d/00nmis.conf /usr/local/nmis8/conf/00nmis.conf.backup #0 8 * * * cat /etc/syslog.conf > /usr/local/nmis8/conf/syslog.conf.backup # These file locations are for Debian/Ubuntu #0 8 * * * cp /etc/apache2/apache2.conf /usr/local/nmis8/conf/httpd.conf.backup #0 8 * * * cp /etc/apache2/sites-available/00nmis.conf /usr/local/nmis8/conf/00nmis.conf.backup #0 8 * * * cat /etc/rsyslog/50-default.conf > /usr/local/nmis8/conf/syslog.conf.backup |
Again, this file was originally generated for Red Hat / CentOS. The lines to change for other distros (e. g. Debian) are the last three. Everything else should be fine.
After you're done, issuing a crontab -l is a good idea to make sure it's all there. Note that crontab -e on most GNU/Linux distros uses the traditional "vi" editor, whereas on Debian, "nano" is used.
Note that there is a reference in this crontab to a <nmisdir>/conf/logrotate.conf file. That file contains a reference to <nmisdir>/logs/ciscopix.log. This is for those who run Cisco PIX/ASA Firewalls. Unfortunately, this file is not in the tarball for NMIS, so you'll get a non-fatal "Hey, there's no ciscopix.log file for me to logrotate!" every day. Fortunately, that's easy to fix.
| Code Block |
|---|
touch <nmisdir>/logs/ciscopix.log |
And now, go ahead and add your hosts to NMIS. There's a nice GUI for doing that which is pretty easy to use. Hey, if an MCSE like me can figure this out, anyone can. :-)
Time to celebrate! You're done! Enjoy your shiny new NMIS installation, and be sure to thank the teams that brought it to you. That would be the NMIS development team and Tobias Oetiker.
SNMPD, Net-SNMP and Testing your Configuration
NMIS8 has been shipped with two nodes configured for collection, to prove it is all working, one is a pingonly node which is ping of the localhost, and the other is localhost using SNMP, and in the <nmisdir>/conf/snmpd.conf is what you need to get it working.
Because you have already installed NET-SNMP earlier, you just need to backup that config and move in the NMIS sample one, you should modify this accordingly to secure the SNMP read access according to your organisations security policy.
| Code Block |
|---|
cd /usr/local/nmis8 mv /etc/snmp/snmpd.conf /etc/snmp/snmpd.conf.orig mv /etc/snmp/snmptrapd.conf /etc/snmp/snmptrapd.conf.orig cp install/snmpd.conf /etc/snmp/snmpd.conf cp install/snmptrapd.conf /etc/snmp cp install/snmptrapd.options /etc/sysconfig/snmptrapd # for CentOS/Red Hat only |
Debian/Ubuntu
The last command in the list above will fail as there is no /etc/sysconfig directory on this platform. Instead the snmptrap service is started by the snmpd service, and is enabled by updating the file /etc/default/snmpd:
| Code Block |
|---|
TRAPDRUN=yes |
Auto Start SNMP Daemons
Ensure that you tell Linux to start SNMPD automatically
| Code Block |
|---|
/sbin/chkconfig snmpd on /sbin/chkconfig snmptrapd on |
Debian/Ubuntu - this is start both snmp and snmptrap services
| Code Block |
|---|
sysv-rc-conf snmpd on service snmpd start |
Start SNMPD Manually
The first time you can start it manually.
| Code Block |
|---|
/etc/init.d/snmpd start /etc/init.d/snmptrapd start |
Run a Test Update
With everything else done, you can run a test update
| Code Block |
|---|
bin/nmis.pl type=update node=localhost debug=true |
The output will look something like
| Code Block |
|---|
NMIS version 8.1.1 09:24:29 runThreads, Starting 09:24:31 runThreads, tables loaded 09:24:31 runThreads, Kernel name of NMIS server is linux 09:24:31 runDaemons, Starting 09:24:31 runDaemons, Finished 09:24:31 runThreads, INFO start of update process 09:24:31 doUpdate, ================================ 09:24:31 doUpdate, Starting, node localhost 09:24:31 init, info of node=localhost loaded 09:24:31 init, info of nmis-system loaded 09:24:31 getNodeCfg, cfg of node=localhost found 09:24:31 init, cfg of node=localhost loaded 09:24:31 loadModel, INFO, model Model loaded 09:24:31 initsnmp, snmp for node=localhost initialized 09:24:31 runPing, Starting localhost (127.0.0.1) with timeout=300 retries=3 packet=56 09:24:34 runPing, localhost is PINGABLE min/avg/max = 0.036/0.070/0.121 ms loss=0% 09:24:34 checkEvent, event localhost-node_down- marked for UP notify and delete 09:24:34 writeHashtoFile, write data to /usr/local/nmis8/var/nmis-event.nmis 09:24:34 getLevelLogEvent, node=localhost, event=Node Up, role=distribution not found in class="event" of model= 09:24:34 runPing, Finished with exit=1 09:24:34 getNodeInfo, Starting 09:24:34 getNodeInfo, sysObjectId=1.3.6.1.4.1.8072.3.2.10, sysObjectName=enterprises.8072.3.2.10 09:24:34 getNodeInfo, sysDescr=Linux nmis-dev 2.6.18-194.el5 #1 SMP Fri Apr 2 14:58:35 EDT 2010 i686 09:24:34 getNodeInfo, oid index 8072, Vendor is net-snmp 09:24:34 selectNodeModel, INFO, Model 'net-snmp' found for Vendor net-snmp 09:24:34 loadModel, INFO, model Model-net-snmp loaded 09:24:34 getDNSloc, Starting 09:24:34 getDNSloc, Finished 09:24:34 checkPIX, Starting 09:24:34 checkPIX, Finished 09:24:34 getNodeInfo, Finished with exit=1 09:24:34 getIntfInfo, Starting 09:24:34 getIntfInfo, Get Interface Info of node localhost, model net-snmp 09:24:34 getIntfInfo, ifIndex=1 ifDescr=lo ifType=softwareLoopback ifAdminStatus=up ifOperStatus=up ifSpeed=10000000 09:24:34 getIntfInfo, ifIndex=2 ifDescr=eth0 ifType=ethernetCsmacd ifAdminStatus=up ifOperStatus=up ifSpeed=1000000000 09:24:34 getIntfInfo, ifIndex=3 ifDescr=sit0 ifType=tunnel ifAdminStatus=down ifOperStatus=down ifSpeed=0 09:24:34 getIntfInfo, Getting Device IP Address Table 09:24:34 getIntfInfo, ifIndex=1, addr=127.0.0.1 mask=255.0.0.0 09:24:34 getIntfInfo, ifIndex=2, addr=192.168.1.7 mask=255.255.255.0 09:24:34 getIntfInfo, ifIndex 1, collect=false, found softwareLoopback in ifType 09:24:34 getIntfInfo, ifIndex 2, collect=true 09:24:34 getIntfInfo, ifIndex 3, collect=false, ifAdminStatus eq down|testing|null 09:24:34 getIntfInfo, Finished 09:24:34 doUpdate, node=localhost role=distribution type=server 09:24:34 doUpdate, vendor=net-snmp model=net-snmp interfaces=3 09:24:34 getEnvInfo, Starting 09:24:34 getEnvInfo, Get Environment Info of node localhost, model net-snmp 09:24:34 getEnvInfo, No class 'environment' declared in Model 09:24:34 getEnvInfo, Finished 09:24:34 getCBQoS, no collecting (none) for node localhost 09:24:34 getCalls, no collecting for node localhost 09:24:34 runReach, Starting node localhost, type=server 09:24:34 runReach, Outage for localhost is 09:24:34 runReach, Getting Interface Utilisation Health 09:24:34 getSummaryStats, Start type=interface, index=2, start=-15 minutes, end=1300973074 09:24:34 getTypeName, ERROR (localhost) type=interface not found in graphtype table 09:24:34 getDBName, ERROR (localhost) database name not found graphtype=interface (section=0), index=2, item= 09:24:34 runReach, Intf Summary in= out= intsumm=200 count=1 09:24:34 runReach, Calculation of health=100 09:24:34 runReach, Reachability and Metric Stats Summary 09:24:34 runReach, collect=true (Node table) 09:24:34 runReach, ping=100 (normalised) 09:24:34 runReach, cpuWeight=100 (normalised) 09:24:34 runReach, memWeight=100 (normalised) 09:24:34 runReach, intWeight=100.00 (100 less the actual total interface utilisation) 09:24:34 runReach, responseWeight=100 (normalised) 09:24:34 runReach, total number of interfaces=2 09:24:34 runReach, total number of interfaces up=0 09:24:34 runReach, total number of interfaces collected=1 09:24:34 runReach, total number of interfaces coll. up= 09:24:34 runReach, availability=100 09:24:34 runReach, cpu= 09:24:34 runReach, health=100 09:24:34 runReach, intfColUp= 09:24:34 runReach, intfCollect=1 09:24:34 runReach, intfTotal=2 09:24:34 runReach, intfUp=0 09:24:34 runReach, loss=0 09:24:34 runReach, mem=100 09:24:34 runReach, operCount= 09:24:34 runReach, operStatus= 09:24:34 runReach, reachability=100 09:24:34 runReach, responsetime=0.070 09:24:34 updateRRD, Starting RRD Update Process, type=health, index=,item= 09:24:34 getFileName, filename of type=health is /usr/local/nmis8/database/health/server/localhost-reach.rrd 09:24:34 updateRRD, file=/usr/local/nmis8/database/health/server/localhost-reach.rrd not found, try nmis4 format 09:24:34 getFileName, filename of type=health is /usr/local/nmis8/database/health/server/127.0.0.1-reach.rrd 09:24:34 createRRD, Starting 09:24:34 createRRD, check and/or create RRD database /usr/local/nmis8/database/health/server/localhost-reach.rrd 09:24:34 optionsRRD, type health 09:24:34 optionsRRD, INFO, using database format 'default' 09:24:34 createRRD, options to create database /usr/local/nmis8/database/health/server/localhost-reach.rrd 09:24:34 createRRD, -b 09:24:34 createRRD, 1300973070 09:24:34 createRRD, -s 09:24:34 createRRD, 300 09:24:34 createRRD, DS:availability:GAUGE:900:0:100 09:24:34 createRRD, DS:health:GAUGE:900:0:100 09:24:34 createRRD, DS:intfColUp:GAUGE:900:0:U 09:24:34 createRRD, DS:intfCollect:GAUGE:900:0:U 09:24:34 createRRD, DS:intfTotal:GAUGE:900:0:U 09:24:34 createRRD, DS:intfUp:GAUGE:900:0:U 09:24:34 createRRD, DS:loss:GAUGE:900:0:100 09:24:34 createRRD, DS:reachability:GAUGE:900:0:100 09:24:34 createRRD, DS:responsetime:GAUGE:900:0:U 09:24:34 createRRD, RRA:AVERAGE:0.5:1:2304 09:24:34 createRRD, RRA:AVERAGE:0.5:6:1536 09:24:34 createRRD, RRA:AVERAGE:0.5:24:2268 09:24:34 createRRD, RRA:AVERAGE:0.5:288:1890 09:24:34 createRRD, RRA:MAX:0.5:1:2304 09:24:34 createRRD, RRA:MAX:0.5:6:1536 09:24:34 createRRD, RRA:MAX:0.5:24:2268 09:24:34 createRRD, RRA:MAX:0.5:288:1890 09:24:34 createRRD, RRA:MIN:0.5:1:2304 09:24:34 createRRD, RRA:MIN:0.5:6:1536 09:24:34 createRRD, RRA:MIN:0.5:24:2268 09:24:34 createRRD, RRA:MIN:0.5:288:1890 09:24:34 createRRD, INFO (localhost) created RRD /usr/local/nmis8/database/health/server/localhost-reach.rrd 09:24:35 createRRD, Finished 09:24:35 updateRRD, DS intfTotal:intfColUp:intfCollect:health:intfUp:availability:reachability:responsetime:loss 09:24:35 updateRRD, value N:2:0:1:100:0:100:100:0.070:0 09:24:35 runReach, Finished 09:24:35 writeHashtoFile, write data to /usr/local/nmis8/var/localhost-view.nmis 09:24:35 writeHashtoFile, write data to /usr/local/nmis8/var/localhost-node.nmis 09:24:35 doUpdate, Finished 09:24:35 runThreads, ### continue normally ### 09:24:35 runThreads, === debug output suppressed with debug=1 === 09:24:35 End of bin/nmis.pl Processed 1 nodes ran for 6 seconds. |
Run a Test Collection
With everything else done, you can run a test update
| Code Block |
|---|
# bin/nmis.pl type=collect node=localhost debug=true |
The output will look something like
| Code Block | ||
|---|---|---|
| ||
NMIS version 8.1.1 09:27:47 runThreads, Starting 09:27:47 runThreads, tables loaded 09:27:47 runThreads, Kernel name of NMIS server is linux 09:27:47 runDaemons, Starting 09:27:47 runDaemons, Finished 09:27:47 doCollect, ================================ 09:27:47 doCollect, Starting, node localhost 09:27:47 init, info of node=localhost loaded 09:27:47 init, info of nmis-system loaded 09:27:47 getNodeCfg, cfg of node=localhost found 09:27:47 init, cfg of node=localhost loaded 09:27:47 loadModel, INFO, model Model-net-snmp loaded 09:27:47 initsnmp, snmp for node=localhost initialized 09:27:47 doCollect, node=localhost role=distribution type=server 09:27:47 doCollect, vendor=net-snmp model=net-snmp interfaces=3 09:27:47 runPing, Starting localhost (127.0.0.1) with timeout=300 retries=3 packet=56 09:27:49 runPing, localhost is PINGABLE min/avg/max = 0.026/0.034/0.041 ms loss=0% 09:27:49 runPing, Finished with exit=1 09:27:49 updateNodeInfo, Starting Update Node Info, node localhost 09:27:49 updateNodeInfo, sysUpTime: Old=0:05:49 New=0:06:03 09:27:49 checkPIX, Starting 09:27:49 checkPIX, Finished 09:27:49 updateNodeInfo, Finished with exit=1 09:27:49 getNodeData, Starting Node get data, node localhost 09:27:49 getValues, no oid loaded for section= 09:27:49 getIntfData, Starting Interface get data, node localhost 09:27:49 getIntfData, lo: ifIndex=1, was => OperStatus=up, ifAdminStatus=up, Collect=false 09:27:49 getIntfData, NOT Collected: lo: ifIndex=1, OperStatus=up, ifAdminStatus=up, Interface Collect=false 09:27:49 getIntfData, eth0: ifIndex=2, was => OperStatus=up, ifAdminStatus=up, Collect=true 09:27:49 getIntfData, collect interface index=2 09:27:49 getIntfData, status now admin=up, oper=up was admin=up, oper=up 09:27:49 getIntfData, rrdData section interface, ds ifOperStatus, value=100, option=gauge,0:100 09:27:49 getIntfData, rrdData section interface, ds ifInOctets, value=14431330, option=counter,0:U 09:27:49 getIntfData, rrdData section interface, ds ifOutOctets, value=707043, option=counter,0:U 09:27:49 updateRRD, Starting RRD Update Process, type=interface, index=2,item= 09:27:49 getFileName, filename of type=interface is /usr/local/nmis8/database/interface/server/localhost/localhost-eth0.rrd 09:27:49 updateRRD, database /usr/local/nmis8/database/interface/server/localhost/localhost-eth0.rrd exists and is R/W 09:27:49 updateRRD, DS ifOperStatus:ifInOctets:ifOutOctets 09:27:49 updateRRD, value N:100:14431330:707043 09:27:49 getSummaryStats, Start type=interface, index=2, start=-6 hours, end=1300973269 09:27:49 getDBName, database name=/usr/local/nmis8/database/interface/server/localhost/localhost-eth0.rrd 09:27:49 getSummaryStats, name=availability, index=2, value=NaN 09:27:49 getSummaryStats, name=inputUtil, index=2, value=NaN 09:27:49 getSummaryStats, name=outputUtil, index=2, value=NaN 09:27:49 getSummaryStats, name=totalUtil, index=2, value=NaN 09:27:49 getSummaryStats, name=inputBits, index=2, value=NaN 09:27:49 getSummaryStats, name=outputBits, index=2, value=NaN 09:27:49 getIntfData, sit0: ifIndex=3, was => OperStatus=down, ifAdminStatus=down, Collect=false 09:27:49 getIntfData, NOT Collected: sit0: ifIndex=3, OperStatus=down, ifAdminStatus=down, Interface Collect=false 09:27:49 getIntfData, : ifIndex=, was => OperStatus=, ifAdminStatus=, Collect= 09:27:49 getIntfData, NOT Collected: : ifIndex=, OperStatus=, ifAdminStatus=, Interface Collect= 09:27:49 getIntfData, : ifIndex=, was => OperStatus=, ifAdminStatus=, Collect= 09:27:49 getIntfData, NOT Collected: : ifIndex=, OperStatus=, ifAdminStatus=, Interface Collect= 09:27:49 getIntfData, : ifIndex=, was => OperStatus=, ifAdminStatus=, Collect= 09:27:49 getIntfData, NOT Collected: : ifIndex=, OperStatus=, ifAdminStatus=, Interface Collect= 09:27:49 getIntfData, : ifIndex=, was => OperStatus=, ifAdminStatus=, Collect= 09:27:49 getIntfData, NOT Collected: : ifIndex=, OperStatus=, ifAdminStatus=, Interface Collect= 09:27:49 getIntfData, Finished 09:27:49 getEnvData, Starting 09:27:49 getEnvData, Get Environment Data of node localhost, model net-snmp 09:27:49 getEnvData, No class 'environment' declared in Model 09:27:49 getEnvData, Finished 09:27:49 getCBQoS, no collecting (none) for node localhost 09:27:49 getCalls, no collecting for node localhost 09:27:49 runServer, Starting server device/storage collection, node localhost 09:27:49 getValues, collect of section device skipped by indexed section 09:27:49 getValues, collect of type cpu skipped by NON indexed section, check this Model 09:27:49 runServer, device Descr=network interface sit0, Type=1.3.6.1.2.1.25.3.1.4 09:27:49 getValues, collect of type cpu skipped by NON indexed section, check this Model 09:27:49 runServer, device Descr=Guessing that there's a floating point co-processor, Type=1.3.6.1.2.1.25.3.1.12 09:27:49 getValues, collect of type cpu skipped by NON indexed section, check this Model 09:27:49 runServer, device Descr=SCSI disk (/dev/sda), Type=1.3.6.1.2.1.25.3.1.6 09:27:49 getValues, collect of type cpu skipped by NON indexed section, check this Model 09:27:49 runServer, device Descr=network interface eth0, Type=1.3.6.1.2.1.25.3.1.4 09:27:49 getValues, collect of type cpu skipped by NON indexed section, check this Model 09:27:49 runServer, device Descr=VMware Virtual IDE CDROM Drive, Type=1.3.6.1.2.1.25.3.1.6 09:27:49 getValues, collect of type cpu skipped by NON indexed section, check this Model 09:27:49 runServer, device Descr=network interface lo, Type=1.3.6.1.2.1.25.3.1.4 09:27:49 getValues, collect of type cpu skipped by NON indexed section, check this Model 09:27:49 runServer, device Descr=Angel___3300_Series@192.168.1.8, Type=1.3.6.1.2.1.25.3.1.5 09:27:49 getValues, collect of type cpu skipped by NON indexed section, check this Model 09:27:49 runServer, device Descr=GenuineIntel: Intel(R) Core(TM)2 Duo CPU T9550 @ 2.66GHz, Type=1.3.6.1.2.1.25.3.1.3 09:27:49 runServer, cpu Load=0.00, Descr=GenuineIntel: Intel(R) Core(TM)2 Duo CPU T9550 @ 2.66GHz 09:27:49 updateRRD, Starting RRD Update Process, type=hrsmpcpu, index=768,item= 09:27:49 getFileName, filename of type=hrsmpcpu is /usr/local/nmis8/database/health/server/localhost-hrsmpcpu768.rrd 09:27:49 updateRRD, database /usr/local/nmis8/database/health/server/localhost-hrsmpcpu768.rrd exists and is R/W 09:27:49 updateRRD, DS hrCpuLoad 09:27:49 updateRRD, value N:3 09:27:49 runServer, storage Type=1.3.6.1.2.1.25.2.1.4, Size=5078687, Used=44473, Units=4096 09:27:49 updateRRD, Starting RRD Update Process, type=hrdisk, index=6,item= 09:27:49 getFileName, filename of type=hrdisk is /usr/local/nmis8/database/health/server/localhost-hrdisk6.rrd 09:27:49 updateRRD, database /usr/local/nmis8/database/health/server/localhost-hrdisk6.rrd exists and is R/W 09:27:49 updateRRD, DS hrDiskSize:hrDiskUsed 09:27:49 updateRRD, value N:20802301952:182161408 09:27:49 runServer, storage Type=1.3.6.1.2.1.25.2.1.3, Size=4194296, Used=0, Units=1024 09:27:49 updateRRD, Starting RRD Update Process, type=hrvmem, index=,item= 09:27:49 getFileName, filename of type=hrvmem is /usr/local/nmis8/database/health/server/localhost-hrvmem.rrd 09:27:49 updateRRD, database /usr/local/nmis8/database/health/server/localhost-hrvmem.rrd exists and is R/W 09:27:49 updateRRD, DS hrVMemUsed:hrVMemSize 09:27:49 updateRRD, value N:0:4294959104 09:27:49 runServer, storage Type=1.3.6.1.2.1.25.2.1.4, Size=101086, Used=12273, Units=1024 09:27:49 runServer, storage Type=1.3.6.1.2.1.25.2.1.2, Size=1035108, Used=968328, Units=1024 09:27:49 updateRRD, Starting RRD Update Process, type=hrmem, index=,item= 09:27:49 getFileName, filename of type=hrmem is /usr/local/nmis8/database/health/server/localhost-hrmem.rrd 09:27:49 updateRRD, database /usr/local/nmis8/database/health/server/localhost-hrmem.rrd exists and is R/W 09:27:49 updateRRD, DS hrMemUsed:hrMemSize 09:27:49 updateRRD, value N:991567872:1059950592 09:27:49 runServer, storage Type=1.3.6.1.2.1.25.2.1.4, Size=0, Used=0, Units=0 09:27:49 runServer, storage Type=1.3.6.1.2.1.25.2.1.4, Size=11649262, Used=876570, Units=4096 09:27:49 updateRRD, Starting RRD Update Process, type=hrdisk, index=4,item= 09:27:49 getFileName, filename of type=hrdisk is /usr/local/nmis8/database/health/server/localhost-hrdisk4.rrd 09:27:49 updateRRD, database /usr/local/nmis8/database/health/server/localhost-hrdisk4.rrd exists and is R/W 09:27:49 updateRRD, DS hrDiskSize:hrDiskUsed 09:27:49 updateRRD, value N:47715377152:3590430720 09:27:49 runServer, storage Type=1.3.6.1.2.1.25.2.1.1, Size=1035108, Used=172032, Units=1024 09:27:49 runServer, storage Type=1.3.6.1.2.1.25.2.1.4, Size=2539342, Used=240465, Units=4096 09:27:49 updateRRD, Starting RRD Update Process, type=hrdisk, index=5,item= 09:27:49 getFileName, filename of type=hrdisk is /usr/local/nmis8/database/health/server/localhost-hrdisk5.rrd 09:27:49 updateRRD, database /usr/local/nmis8/database/health/server/localhost-hrdisk5.rrd exists and is R/W 09:27:49 updateRRD, DS hrDiskSize:hrDiskUsed 09:27:49 updateRRD, value N:10401144832:984944640 09:27:49 runServer, Finished 09:27:49 runServices, Starting Services stats, node=localhost, nodeType=server |