evolution creations

So after coming back from holiday’s the only box that was on my network was my Linux router/server/firewall. In powering up all the other Windows machines two of them came back online ok and started reporting graphs, but my main Windows box didn’t. The error when I went into Cacti Admin > Console > Devices > The Device was: SNMP Error right under the upper left corner of the window under SNMP Information.

The first thing that I went to check was under Control Panels > Administrative Tools > Services and verified that the SNMP Service and SNMP Trap were both set to automatically start, and are both running. I went in to check the configuration of the SNMP service and the IP address that it was set to report to is correct and the community is fine.

So to break through all the walls here, the first test that you should run is:

snmpwalk -Os -c public -v 1 xxx.xxx.xxx.xx system

Replace the XXX’s with the actual internal IP address. Now what’s interesting for me is that my main box has two NIC’s on it; 192.168.1.8 and 192.168.1.20. When I ran the test against the first IP address, I received the below error:

george@otani:~$ snmpwalk -Os -c public -v 1 192.168.1.8 system
Timeout: No Response from 192.168.1.8

But when I ran it against the 192.168.1.20 IP, I received a response:

george@otani:~$ snmpwalk -Os -c public -v 1 192.168.1.20 system
sysDescr.0 = STRING: Hardware: x86 Family 6 Model 15 Stepping 11 AT/AT COMPATIBLE – Software: Windows Version 6.0 (Build 6001 Multiprocessor Free)
sysObjectID.0 = OID: enterprises.311.1.1.3.1.1
sysUpTimeInstance = Timeticks: (216315) 0:36:03.15
sysContact.0 = STRING:
sysName.0 = STRING: COMPUTERNAME
sysLocation.0 = STRING:
sysServices.0 = INTEGER: 76

This boils down to a firewall or some sort of block that is causing this error. If you don’t have two NIC’s, start with troubleshooting it at the local firewall level or at your router level. But changing the device IP from 192.168.1.8 to 192.168.1.20, I received the following within Cacti:

GEORGE-INTEL-DESKTOP (192.168.1.20) SNMP Information
System: Hardware: x86 Family 6 Model 15 Stepping 11 AT/AT COMPATIBLE – Software: Windows Version 6.0 (Build 6001 Multiprocessor Free)
Uptime: 218222 (0 days, 0 hours, 36 minutes)
Hostname: COMPUTERNAME
Location:
Contact:

On the new Apache2 instance that I have running Cacti, I ran into a problem where I get the error message:

apache2: Could not reliably determine the server’s fully qualified domain name, using 127.0.1.1 for ServerName

To fix this, I have to add the following line into the /etc/apache2/httpd.conf file:

ServerName nameofserver

Replace nameofserver with the actual name of the server that you want to have Apache configured under.

So to jazz things up again, I wanted to move away from the PHP poller, and install Spined, which is the poller daemon that can be used instead of the PHP program and is much more efficient for larger systems.

To start with, you want to download the latest version of Spine from the website here. Once you have downloaded the tar ball (I downloaded it to my /home/george directory (you will want to change george to whatever your current user is)), and run:

tar zxvf cacti-spine-0.8.7a.tar.gz

and then you want to:

cd cacti-spine-0.8.7a

From this directory, you will need to run:

./configure

Now when I did this, I got into loads of problems, because it was telling me that:

error: cannot run /bin/bash conifg/config.sub

To get around this, I had to run:

sudo apt-get install libmysqlclient15-dev libsnmp9-dev automake libtool

Then I was able to run:

aclocal
libtoolize –force
autoconf
autoheader
automake
./configure

In that order. Each of the above lines needs to be run separetely. Once you successfully run the ./configure, you can run:

make

to make the binary. Now once that has gone through for you, you will need to create a directory:

sudo mkdir /usr/local/spine

and then execute:

sudo cp spine* /usr/local/spine

This will move the spine binary and spine.conf files from your /home directory to the /usr/local/spine directory.

Then you want to run:

sudo joe /usr/local/spine/spine.conf

and modify the configuration to the proper settings for your cacti database instance. If you used my previous guide your information should be:

DB_Host localhost
DB_Database cacti
DB_User cacti
DB_Pass xxxxxx
DB_Port 3306

You will replace the xxxxxx with the password that you setup for your Cacti database during the installation.

Finally you will need to login to Cacti, and go to Console > Settings > Paths, and modify the path for Spine under the Alternative Poller Path section:

/usr/local/spine/spine

When you click on save, the words under the path field for Spine should read:

[OK: FILE FOUND]

Then you will want to click on the Poller tab and set this from cmd.php to spine and save your settings. This should be it and you should be using Spine as your poller!

Reference 1
Reference 2

So I’ve had the WORST luck in trying to figure out how to setup Cacti on Ubuntu so it will properly poll my system and give me graphs. Part of the reasoning behind this is that I’m in the process of replacing my existing Netgear router with a Ubuntu box (guide to come shortly!) but a prereq of doing this is to setup reporting on the Ubuntu box so I can see all kinds of metrics on disk, IO, network bandwidth, etc–all the fun stuff that makes you jump up and down with. My previous attempts went through fine, but the problem was that I didn’t seem to be able to get stats polled from my devices, namely network interfaces.

To begin with always make sure your system is up2date:

sudo apt-get update
sudo apt-get upgrade

Now that we have that out of the way, make sure your system has SNMPD installed:

sudo apt-get install snmp snmpd

Now that it is installed, we need to do a bit of configuration to it. Run:

sudo nano /etc/default/snmpd

and you’ll see something like:

# This file controls the activity of snmpd and snmptrapd

# MIB directories. /usr/share/snmp/mibs is the default, but
# including it here avoids some strange problems.
export MIBDIRS=/usr/share/snmp/mibs

# snmpd control (yes means start daemon).
SNMPDRUN=yes

# snmpd options (use syslog, close stdin/out/err).
SNMPDOPTS=’-Lsd -Lf /dev/null -u snmp -I -smux -p /var/run/snmpd.pid’

# snmptrapd control (yes means start daemon). As of net-snmp version
# 5.0, master agentx support must be enabled in snmpd before snmptrapd
# can be run. See snmpd.conf(5) for how to do this.
TRAPDRUN=no

# snmptrapd options (use syslog).
TRAPDOPTS=’-Lsd -p /var/run/snmptrapd.pid’

# create symlink on Debian legacy location to official RFC path
SNMPDCOMPAT=yes

Make sure that you change:

SNMPDRUN from No to Yes, and you remove the 127.0.0.1 from the SNMPDOPTS line. You’ll see what I’m saying when you compare yours to mine that you see above that I’ve made bold.

Then we are going to configure the communities by running:

sudo nano /etc/snmp/snmpd.conf

Do a search within the conf file and you should see:

# sec.name source community
#com2sec paranoid default public
com2sec readonly localhost public
#com2sec readwrite default private

By default the first line is the one that is commented and the last two are uncommented. I mearly commented out the first line (its the one in italics) and removed the comment from the second line. Also I changed the source from default to localhost and left the community name as public.

Now comes the test, you want to make sure that snmp is up and running. Since changes were made to the configuration you want to give those services a swift kick in the pants:

sudo /etc/init.d/snmpd restart

Once that is done you’ll want to run:

snmpwalk -Os -c public -v 1 localhost system

You should get a mess of information returned:

sysDescr.0 = STRING: Linux otani 2.6.24-19-generic #1 SMP Wed Jun 18 14:43:41 UTC 2008 i686
sysObjectID.0 = OID: netSnmpAgentOIDs.10
sysUpTimeInstance = Timeticks: (227377) 0:37:53.77
sysContact.0 = STRING: Root (configure /etc/snmp/snmpd.local.conf)
sysName.0 = STRING: otani
sysLocation.0 = STRING: Unknown (configure /etc/snmp/snmpd.local.conf)
sysORLastChange.0 = Timeticks: (0) 0:00:00.00
sysORID.1 = OID: snmpFrameworkMIBCompliance
sysORID.2 = OID: snmpMPDCompliance
sysORID.3 = OID: usmMIBCompliance
sysORID.4 = OID: snmpMIB
sysORID.5 = OID: tcpMIB
sysORID.6 = OID: ip
sysORID.7 = OID: udpMIB
sysORID.8 = OID: vacmBasicGroup
sysORDescr.1 = STRING: The SNMP Management Architecture MIB.
sysORDescr.2 = STRING: The MIB for Message Processing and Dispatching.
sysORDescr.3 = STRING: The management information definitions for the SNMP User-based Security Model.
sysORDescr.4 = STRING: The MIB module for SNMPv2 entities
sysORDescr.5 = STRING: The MIB module for managing TCP implementations
sysORDescr.6 = STRING: The MIB module for managing IP and ICMP implementations
sysORDescr.7 = STRING: The MIB module for managing UDP implementations
sysORDescr.8 = STRING: View-based Access Control Model for SNMP.
sysORUpTime.1 = Timeticks: (0) 0:00:00.00
sysORUpTime.2 = Timeticks: (0) 0:00:00.00
sysORUpTime.3 = Timeticks: (0) 0:00:00.00
sysORUpTime.4 = Timeticks: (0) 0:00:00.00
sysORUpTime.5 = Timeticks: (0) 0:00:00.00
sysORUpTime.6 = Timeticks: (0) 0:00:00.00
sysORUpTime.7 = Timeticks: (0) 0:00:00.00
sysORUpTime.8 = Timeticks: (0) 0:00:00.00

If you get something else like:

Timout: No response from Localhost

Go back and make sure you’ve setup your communities properly and restart SNMPD from the /etc/init.d/snmpd restart command.

Next you are going to want to setup your web and MySQL server’s. If you’ve already done this great, but for those who are starting on a vanilla installation, go ahead and run the following:

sudo apt-get install mysql-server apache2 libapache2-mod-php5 php5-mysql php5-cli php5-snmp

You’ll need to test that Apache and PHP are up and running, so run:

sudo joe /var/www/phpinfo.php

and type in the phpinfo bit (unfortunately it won’t properly display through Blogger if I type in the code, but you can always find it here on how to create a test page).

Save your file and go to: http://localhost/phpinfo.php and if you get the PHP info page, you are all set there!

Next you’ll want to install phpmyadmin. It’s not a requirement, but I always do this so you have a graphical interface to manage your MySQL instance:

sudo apt-get install phpmyadmin

And you can test that this is running by going to: http://localhost/phpmyadmin and logging in with root as the username and the password you setup during the installation process of MySQL above.

Now that we have that squared away, we want to install Cacti using:

sudo apt-get install cacti

The installation process is fairly straight forward. You will be asked:

1. Configure libphp-adodb = Press OK
2. Configure database for cacti with dbconfig-common = Press OK
3. Password for MySQL database Administrator = Type in the password that you configured in the MySQL installation process and the one that you verified works through phpMyAdmin above.
4. MySQL Application password for Cacti = This is the password for the new Cacti database instance that is going to be created. Make this a good one, as it’s not going to be the same one that is used to log into Cacti to administer it.
5. Webserver Type = apache2

So now that cacti has been installed and we are trucking right along, we’ll want to pull up the web installer: http://localhost/cacti. You have to go through the web-based installer process:

1. Click Next on the first screen
2. Select NEW INSTALL and check that the information is correct
3. Make sure that everything shows up as [FOUND] and resolve any conflicts (there shouldn’t be as in my installation I didn’t have to do anything else).
4. Click Finish

When you get to the login page, your default credentials on the first run are:

• Username = admin
• Password = admin

Obviously be a good boy and on the next screen make sure you change this to something proper. This will be the admin password that you change, so make it a good one!

Now comes the fun part, you are going to setup graphs! But before you can get to that you have to understand the process by which you set things up:

1. Setup Devices
2. Create graphs for devices
3. Optional – Create a new tree for the devices
4. Associate Devices to a Tree

This is the part that I got stumbled up on for awhile as it’s not the most intuitive system to go through. So to start with click on Console from the upper right corner, and click on the link labeled Create devices for network.

There should already be a device there, but go ahead and delete it and create a new one from scratch. (The Add button to add a new one is in the upper right hand corner of the frame). In the add menu, there will be several fields that you need to fill in:

• Description – name this something that you will recognize, for me it was localhost-2
• Hostname – 127.0.0.1
• Host Template – ucd/net SNMP Host
• Description – Whatever you want to put there to make it so you know what this is
• Down Device Detection – Ping
• Ping Method – UDP Ping
• Ping Port – 23
• Ping timout value – 400
• Ping retry count – 1
• SNMP Version – 1

When you select SNMP version to 1, additional options will open up for you:

• SNMP Community – public, or whatever name you specified in your SNMP community that was configured earlier in this guide.
• SNMP Port – 161
• SNMP Timeout – 500
• Maximum OID’s Per Get Request – 10

Once you have that click on Create to create your device. After you click on create, the device will refresh with additional options at the bottom for:

• Associated Graph Templates
• Associated Data Source Templates

If you scroll to the top of the frame, you’ll see a link labeled Create Graphs for this Host. Click on the link to get you to the next screen. Under Graph Templates, select the check boxes next to:

• CPU, Load Average, and Memory

Under Data Query [SNMP - Interface Statistics], click on the drop down menu and select In/Out Bits with Total Bandwidth and click the Create button.

We’re almost there. We have to either first create a tree to associate this new device to, or we have to associate this device to an existing tree. By default, Cacti will install with a Default Tree. And for those like me, who are setting this up on one box, and this will be the only box that Cacti monitors, we can skip creating a new tree and go straight to associating this to an existing tree. If you want to create your own tree, it’s simply a matter of clicking on the Graph Trees link on the left menu, and adding a new tree with a new name.

To associate your device with a tree, it’s simply just clicking on the Devices link on the left menu, and clicking the checkbox next to your device, and from the drop down menu select Place on Tree (Default Tree) and click on the Go button. You’ll be asked to confirm, just confirm with default values and that’s it!

If everything went well within a few minutes you should have some graphs being generated!

Reference

So my one Ubuntu server box after the last server update died and never came back to life. It was on an old IBM Intellistation P4 box, and the POST beep codes tell me that it’s either the system RAM or the system board. Since neither I’m in a position where I’m about to pour more cash into, I decided to just put a Ubuntu VM on my main box and go from there. After all the main box these days seems to be just idling away, so why not!

The problem was that I couldn’t remember how to mount a network drive, so it booted up at startup. After a few minutes of souring and trying a few things, came across the answer.

First run the following command:

sudo apt-get install smbfs

This will install the Samba File System, but more importantly it will also install the smbmount application.

Test to make sure that you can from the machine that you are setting up the mount point reach the network file system. You can do this via the command:

sudo smbmount //servername/directory /mountpoint -o username=putusernamehere,password=putpasswordhere

Obviously replace the necessary values above with the real information. Mine looks like:

sudo smbmount //192.168.1.7/public /storage -o username=george,password=xxxxxxxxx

Doing this, I received the error message:

mount error: can not change directory into mount target /storage

This is resolved by simply running the command:

sudo mkdir /storage

To add this mount point to your startup you will first backup your fstab file:

sudo cp /etc/fstab /etc/fstab_backup

once that is backed up, you will:

sudo nano /etc/fstab

Add in the following line:

//servername/directory /mountpoint smbfs username=putusernamehere,password=putpasswordhere 0 0

Mine looks like:

//192.168.1.7/public /storage smbfs username=george,password=xxxxxxxxx 0 0

Then writeout the file and exit nano, and run:

sudo mount -a

This will reload the mount points and if successful you will have full access to your network drive via the /storage directory!

Some distributions run a caching nameserver in the background out of the box while some do not. To clear the cache if you have such a daemon running, simply restart the nscd service in ubuntu.

Most DNS clients will cache the results of name resolution request to speed up multiple lookups to the same URL. Just think about how many requests are made to the same domain when visiting a single web page. Every file, image, style sheet, etc. that is on that page and served from the same domain requires a DNS lookup.

So if you have an invalid DNS entry cached on your local client you’ll need to flush it out of the cache so your client can do a new lookup and get the correct information. Or your other option is to wait until that DNS entry expires and the cache flushes it automatically… which typically takes about 24 hours.

In ubuntu if you want to flush DNS cache you need to restart nscd daemon

Install nscd using the following command

sudo aptitude install nscd

Flush DNS Cache in Ubuntu Using the following command

sudo /etc/init.d/nscd restart

Reference