Autodiscovery

Autodiscovery

Autodiscovery is a tool that can help you find hosts in your network that you want to import into Monitor. Autodiscovery is found under the Manage menu in OP5 Monitor.

Discovery jobs

To start discovering hosts in your network you need a discovery job that includes the details for how the discovery will be made.

All available discovery jobs are listed on the Run a discovery tab.

Create new jobs

To create a new discovery job, click the Create new job button on the Run a discovery tab. This will bring up a dialog where you can fill out the details of the new job.

Discovery methods

Autodiscovery currently supports three different methods for exploring a network: Seed node, IP range and IP list.

Seed node method

The seed node method works by giving an IP address from where Autodiscovery will start asking for its neighbours. The process is repeated recursively so that all the neighbours are also asked for their neighbours, until no further neighbours can be found. This will result in a network of linked hosts.
The seed node method uses the LLDP and ARP protocol via the SNMP protocol to query the discovered hosts about their neighbouring information. ARP is an optional protocol which can be configured. Please check the configuration options for more details.

IP range method

The IP range method works by giving an IP range and sending ICMP echo requests (also known as ping) to each address in the range. This will result in a list of hosts that reply to the ICMP request.
The hosts can be queried for further information via SNMP using the Device identification option, which is explained further below.

IP range syntax

Autodiscovery uses the Nmap utility for ICMP based discovery and supports the same syntax for specifying IP ranges. You can use both CIDR-style addressing like 192.168.10.0/24 or octet range addressing like 192.168.10.0-255, 192.168.10-20.0-255 or 192.168.10,11,12.0-255. See the Nmap documentation for more details.

IP list method

The IP list method works by giving a list of IP addresses or ranges that is each processed as the IP range method. The items can be separated by newline or space characters.

Device identification

Device identification is an option that can be enabled for the IP Range and IP list methods. (The Seed node method will automatically use device identification since it is already using SNMP to do the discovery.) This will attempt to fetch more information about the discovered hosts via the SNMP protocol.
Note that using this option can result in additional hosts being discovered, if the initially discovered hosts have relations that were not initially discovered. All discovered hosts will however always be within the initially given IP range or IP list, and hosts outside those parameters are discarded.

Deleting jobs

Discovery jobs can be deleted by clicking the Delete button in the job details, or in the corresponding job's Actions menu on the Run a discovery tab.

Note: Deleting a job will not delete any discoveries already created for that job, but will delete the information about what parameters those executions were run with.

Discoveries

Discoveries are created when you run a discovery job and the actual network discovery is started. You can start a discovery by clicking the Run button in the the job details, or in the corresponding job's Actions menu on the Run a discovery tab.
All available discoveries are listed chronologically with their current status on the View discoveries tab.

Starting discoveries

To start a new execution of a discovery job first find the job you want to run on the Run a discovery tab, then either select Run from that job's Actions menu, or click on the job name and then click on the Run button in the job dialog.

Stopping discoveries

Running discoveries can be stopped before they finish by clicking the Stop discovery button in the corresponding discovery's Actions menu on the View discoveries tab.

Discovery results

The results of a discovery can be seen by clicking the View results button in the corresponding discovery's Actions menu on the View discoveries tab. The results of a discovery will be a list of discovered IP addresses with additional information about system name, system description. For Seed node discoveries the list will also include information about the parent and child relationships found for each host.

  • System name is primarily fetched from the SNMP object sysName, if device identification is enabled. Otherwise Autodiscovery will attempt to find the hostname by doing a reverse DNS lookup. Reverse DNS lookups can be disabled as part of the service configuration through the no_dns option.
  • System description is fetched from the SNMP object sysDesc, if device identification is enabled.
  • Parents and Children are the neighbouring hosts found during a Seed node discovery, based on the LLDP and ARP information that is available via SNMP. These are the hosts that are directly connected to the discovered host.

Importing discoveries

The results of a completed discovery can be imported into Monitor by clicking the Import button on the discovery's results page. This brings up a dialog with all hosts that will be imported. Confirm the import by clicking the Send to configuration button. The hosts are now imported to Monitor and the progress is tracked by a progress bar. If the import is successful you can continue to the Monitor configuration page to save the changes.

Selecting hosts for import

By unchecking the check boxes in the list of discovery results it is possible to exclude discovered hosts from being imported. The check box in the list header can be used to select or de-select all discovered hosts.

Free device selection

If the Seed node method was used, unchecking a host will automatically also uncheck all other hosts that would no longer be discoverable without going through an unchecked host. This ensures that all imported hosts are reachable from the seed node. This behavior can be disabled by checking the Free device selection option.

Deleting discoveries

Discoveries can be deleted by clicking the Delete discovery button in the corresponding discovery's Actions menu on the View discoveries tab.

Note:

  • Deleting a discovery is not possible while it is running. It must then be stopped first.
  • Deleting a discovery will not delete the discovery job that was used to start it.
  • Deleting a discovery will not make any changes to your Monitor configuration.

Configuration options

The Autodiscovery service is named Magellan, and its configuration file is found at /opt/magellan/magellan.conf .

Below are some of the options you may want to change:

no_arp

During the discovery process, LLDP and ARP information will be looked up. The ARP table lookup is optional and enabled by default. If a user wish to disable ARP lookup, the configuration option no_arp can be changed to True.

port

To run Magellan on a different port than the default 8008 you can change it here. Remember to change the port also in the HTTP API configuration, or the Autodiscovery UI will no longer be able to talk to Magellan.

Known limitations

  • Autodiscovery does not work across distributed Monitor nodes. It will only be able to find hosts that are reachable from the node it is running on.