Data Collection via SNMP
Data collection via SNMP is a useful alternative for collecting network statistics in systems where Juniper Telemetry Interface (JTI) is not available or in multi-vendor systems. Data collection via SNMP enables the following performance management features:
Collection of interface statistics using SNMP collection tasks that poll the SNMP MIB (Juniper Networks and Cisco devices).
Collection of LSP statistics using SNMP collection tasks that poll the SNMP MIB (Juniper Networks and Cisco devices).
Cisco LSP statistics can also be collected by polling the interface MIB because in Cisco devices, an LSP tunnel is a special interface entry.
Collection of P2MP LSP statistics by polling the Juniper LSP MIB for Juniper Networks devices, or by polling the standard IFMIB for Cisco devices. Even older Juniper devices are supported.
Collection of class of service (CoS) statistics. To collect this data for Juniper Networks devices, the SNMP collector polls the JUNIPER-COS-MIB.
Table 1: OIDs for Interface and LSP Statistics
Table 2: OIDs for CoS Statistics - Juniper Devices
Table 3: OIDs for CoS Statistics - Cisco Devices
CISCO-CLASS-BASED-QOS-MIB:: cbQosClassMapStats. cbQosCMDropByte64
The process involves the following tasks:
Installation of Collectors
The collectors are installed in the same machine as the NorthStar Controller application server (single-server deployment) by the install.sh script when you install the controller itself. Once installed, you can see the collector group of processes:
[root@pcs-q-pod05 ~]# supervisorctl status
analytics:elasticsearch RUNNING pid 3374, uptime 6:33:42 analytics:esauthproxy RUNNING pid 3373, uptime 6:33:42 analytics:logstash RUNNING pid 5600, uptime 6:31:15 collector:es_publisher RUNNING pid 12899, uptime 0:37:03 collector:task_scheduler RUNNING pid 12900, uptime 0:37:03 collector:worker1 RUNNING pid 3385, uptime 6:33:42 collector:worker2 RUNNING pid 3387, uptime 6:33:42 collector:worker3 RUNNING pid 3386, uptime 6:33:42 collector:worker4 RUNNING pid 3388, uptime 6:33:42
Configure Devices in Device Profile and Test Connectivity
Before you can run SNMP collection, you must configure login credentials and SNMP parameters for the devices. In the web UI, from the More Options menu, navigate to Administration > Device Profile. Select a device and click Modify. Click the Access Parameters tab to enter login credentials and the SNMP Parameters tab to enter SNMP parameters.
See Device Profile for detailed instructions on setting up devices with SNMP parameters, and also on testing SNMP connectivity to those devices.
Run Netconf Device Collection
You must run Netconf device collection before attempting to run SNMP traffic collection. This is necessary to establish the baseline network information including the interfaces and LSPs. Once Netconf device collection has been run, SNMP traffic collection tasks have the information they need to poll the interfaces and the LSPs.
Schedule and Run SNMP Data Collection Tasks
Completion of device profiles (Administration > Device Profile) and running Netconf device collection are prerequisites for successfully running SNMP collection.
To schedule a new SNMP collection task, navigate to Administration > Device Collection from the More Options menu.
- Click Add in the upper right corner. The Create New Task window is displayed as shown in Figure 1.
- Enter a name for the task and use the drop-down menu to
select the task type as SNMP Traffic Collection. Click Next.
The next window displayed does not offer any options because at this time, liveNetwork is the only device profile available. Figure 2 shows this window for SNMP traffic collection.
- Click Next to proceed to the
scheduling parameters. The Create New Task - Schedule window is displayed
as shown in Figure 3. At least two
collections are necessary for the calculation of statistics. We recommend
setting up automatic recurrence of the task every 10 to 20 minutes.
Instead of scheduling recurrence, you can select to chain the task after an already-scheduled recurring task, so it launches as soon as the other task completes. When you select the “Chain after another task” radio button, a drop-down list of recurring tasks is displayed from which to select.
- Click Submit to complete the
addition of the new collection task and add it to the Task List. Click
a completed task in the list to display the results in the lower portion
of the window. There are three tabs in the results window: Summary,
Status, and History. An example of the Summary tab is shown in Figure 4. An example of the Status tab
is shown in Figure 5.
You can have only one SNMP traffic collection task per NorthStar server. If you attempt to add a second, the system will prompt you to approve overwriting the first one.
By default, NorthStar only collects statistics from the following interfaces when running SNMP traffic collection:
Physical, logical loopback, or logical management interfaces that can be associated with nodes in NorthStar
Logical interfaces associated with links in NorthStar
Logical interfaces belonging to a VRF
The interface types that can be discovered on devices and that should be used by traffic collection can be modified by editing the /opt/northstar/data/northstar.cfg file. Use a text editing tool such as vi, and use a comma as a separator. For example:
configServer_include_interfaceType=physical, loopbackMgmt, vrfInterface, linksInterface
The supported interface types are:
physical: Physical interfaces, expressed as the interface name without a dot (.) in it.
loopbackMgmt: Loopback and management interfaces expressed as the interface name starting with lo, fxp, me, or em.
vrfIf: Interfaces with which a VRF is associated.
linksIf: Interfaces on links.
all: All interfaces
These supported interface types are also commented in the northstar.cfg file.
Access the Data from the NorthStar Planner
You can access the collected data from the NorthStar Planner for planning and simulation purposes. In the NorthStar Planner, navigate to Traffic > Traffic aggregation. You can aggregate the traffic by hour and create a 24-hour traffic load file for each hour, aggregating the data for that particular hour across multiple days. The resulting file can be used as input into the traffic matrix solver.