Sometimes, NNMi may have difficulties discovering some
nodes. For example, nodes that are far away from NNMi in terms of
number of hops. In such cases, we may want or even need to boost a
discovery, forcing NNMi to load some devices automatically. This can be
done using seeds. This section describes how to work with seeds.
A seed is a list of nodes which has to be loaded into the NNMi
topology. Technically, it's a simple part of NNMi—the administrator
lists IP addresses or hostnames, which have to be forced into NNMi's
topology. But on other hand, it's very important to understand its
meaning and influence on the whole deployment and operation process.
Seeds are very handy when devices cannot be discovered by auto
discovery rules.
How can discovery be boosted with seeds?
For NNM 7.x and earlier the versions, seeds were text files, which
were read by netmon process. NNMi needs to have seeds loaded into a
database and there are two ways to do so:
- Entering manually using NNMi GUI
- Loading from seed file using
nnmloadseeds.ovpl
command
The following is the syntax for these files:
That is:
Number sign (#) is used as a comment sign and anything after it in the line are taken as Notes (see the screenshot in the section Example 2: Create a seedfile for batch load) in Discovery Seed configuration tab in Discovery Configuration. It is always a good practice to keep notes about the nodes you are adding, so the next time you open a Discovery Seed
configuration window you can remember why each of the nodes were added.
It is even more important to keep notes if you have more than one NNMi
administrator, so that each of them can know what is going on in the
network management. The IP address found in the seed is determined as a
primary management address. If it has a hostname, then NNMi takes the
IP address that is resolved by DNS.
Note
For Cisco devices, the loopback address should be used for
discovery. Make sure that the DNS is correctly configured to resolve
the name by loopback addresses.
It is good practice to have one seed file. If you add new devices
into the seed file, they will be loaded into NNMi topology. If you
delete any node from the seed file, they will not be deleted
automatically from NNMi topology and vice versa, if you delete a node
from topology, it will not be deleted from the seed and it will be
discovered again.
The following screenshot displays discovery configuration by providing discovery seeds (Discovery Seeds tab on Discovery configuration window).
Only the name or IP address is required in seeds. No subnet masks are required.
How to load seeds in NNMi
There are two ways in which you can seed nodes into NNMi discovery:
- Adding nodes using the seed configuration tab window
- Creating a seedfile and load it using the
nnmloadseeds.ovpl
command
Let's examine one of these two options using an example.
Example 1: Adding nodes using seed configuration tab window
We have two nodes with IP addresses 10.10.1.1 and 192.168.1.1 and we
will seed them using configuration window. Complete the following steps:
- Select Configuration workspace.
- Select Discovery Configuration view.
- Select Discovery Seeds tab.
- Select the New button in the top left corner of the Discovery Seeds tab.
- Enter
10.10.1.1
in Host Name/IP field and click Save and Close:
- Repeat step 4 and 5 for the IP address 192.168.1.1.
Congratulations! You have loaded two nodes into the seed window,
which you can see in the previous figure. Your nodes are being
discovered now and in a couple of seconds, you will see all the
information regarding these nodes.
Example 2: Create a seedfile for batch load
Now, using batch load command (nnmloadseeds.ovpl
), let's load two more devices, which have IP addresses 10.10.1.2 and 192.168.1.2. Complete the following steps to load nodes using batch command:
- Create a file with a list of IP addresses and save it (we chose filename
/tmp/seedfile
in our example), as follows:
Text after number sign (#) will go as comments in Discovery Seed Results window.
- Run
nnmloadseeds.ovpl
command with file pass parameter as follows:
You should get a response about seed results. In our example, two
seeds were added and no duplicate or invalid seeds were provided
(output example is shown as follows):
- Make sure that nodes were added. Open seed configuration window by selecting Configuration | Discovery Configuration | Discovery Seeds:
Note
As you see in the preceding figure, nodes from seedfile
were loaded with comments provided in a file into Notes field.