...

NMIS talks to 1000's of different vendor products, and some of those devices have less than well performing SNMP agents.  When you have a slow SNMP agent, along with a high interface count, the result can be very slow polling.  For example we have found some SNMP agents which have 1000's of interfaces and an NMIS update takes less than 30 seconds, while other devices have taken over 60 minutes to run an update.

Which Nodes are Taking a Long Time to Poll

In NMIS 8.5.10 a feature was added for granular polling timers, this can be enabled with the NMIS configuration setting log_polling_time, set this value to true to see output like the sample below, the "doCollect" indicates it is from a collect (poll) cycle, the codename and then the model being used.  This should assist to find which nodes are slowing down your poll cycle.

24-Aug-2015 17:55:07,nmis.pl::doCollect#976<br>Poll Time: cloud, PingOnly, 0.01
24-Aug-2015 17:55:07,nmis.pl::doCollect#976<br>Poll Time: gate, PingOnly, 0.03
24-Aug-2015 17:55:07,nmis.pl::doCollect#976<br>Poll Time: golden, PingOnly, 0.02
24-Aug-2015 17:55:07,nmis.pl::doCollect#976<br>Poll Time: bones, net-snmp, 0.31
24-Aug-2015 17:55:07,nmis.pl::doCollect#976<br>Poll Time: excalibur, ESXi, 0.26
24-Aug-2015 17:55:07,nmis.pl::doCollect#976<br>Poll Time: elli, Windows2008, 0.32
24-Aug-2015 17:55:08,nmis.pl::doCollect#976<br>Poll Time: meatball, CiscoRouter, 0.63
24-Aug-2015 17:55:08,nmis.pl::doCollect#976<br>Poll Time: gdc-sw1, Netgear-GS724T, 0.86
24-Aug-2015 17:55:08,nmis.pl::doCollect#976<br>Poll Time: omed, net-snmp, 0.27
24-Aug-2015 17:55:08,nmis.pl::doCollect#976<br>Poll Time: ran, net-snmp, 0.69
24-Aug-2015 17:55:09,nmis.pl::doCollect#976<br>Poll Time: NETGEAR-MANUAL, Netgear-Manual, 2.18
24-Aug-2015 17:55:09,nmis.pl::doCollect#976<br>Poll Time: wanedge1, CiscoRouter, 0.42
24-Aug-2015 17:55:10,nmis.pl::doCollect#976<br>Poll Time: yang, QNAP, 0.87
24-Aug-2015 17:55:10,nmis.pl::doCollect#976<br>Poll Time: kaos, Windows2012, 3.24

NMIS Configuration Options and Features

interface_max_number

NMIS can handle devices with very high device counts, but we have noticed that people don't really know that they have these devices or how many interfaces they have.  Because managing 1000's of interfaces could fill the disk or prevent effective polling, the interface_max_number configuration option was added in NMIS 8.5.8G, the default setting is 5000.

This means that if NMIS finds a device with an interface count over the configured setting, it will NOT collect interface data from that node at all.  This means that the NMIS administrator has to consciously enable handling of devices which high interface counts and they can monitor how well the device and NMIS are handling collecting the data.

...

Polling Locks

This configuration option feature was added in NMIS 8.5.10.  The SNMP variable ifTableLastChange holds the value of sysUpTime when the interface table last changed, that is when an interface was added or deleted.  The default value for this is false.  When set to true the NMIS update process will check this value for each node and only run a complete update of the interface table when this value has changed.  For a node with high interface counts, this can result in not needing to run an update very often at all.  Which tends to work well as these devices do not often have new interface being added. Polling locks work that when an update or collect poll runs on a node, a lock is created, preventing another NMIS process from starting an update or collect.  This is done because an update on a node with high interface counts can go for quite a while, we don't want another update running on the node at the same time.  We also don't want NMIS running a collect on a node which will start an update if an update has never been run, and the server might get caught with too many blocked processes.

NMIS Modelling Options

The following modelling options were introduced in NMIS 8.5.10 to assist with managing devices with very high interface counts.  We have found that the majority of devices do not require these but for some vendors using SNMPv1 or with very high interface counts, these features can improve polling performance dramatically.

A key concept with these features is to avoid running an update cycle for interfaces when changes are detected and to minimise the number of bulk walks NMIS does, most devices handle this no problem, but when there are slow SNMP engines or VERY high interface counts, optimisation is needed.

Below is the sample model which can be added to an NMIS model to use these feature, this section will describe each of the features in more detail the settings below are considered the defaults, as if the features were not present.

Default Interface Handling Options

  'custom' => {
    'interface' => {
      'ifAdminStatus' => 'true',
      'ifNumber' => 'true',
      'ifLastChange' => 'false',
      'ifTableLastChange' => 'false',
      'skipIpAddressTableOnSingle' => 'false',
    }
  },

Suggested Interface Handling for Devices with Interface Polling Issues

  'custom' => {
    'interface' => {
      'ifAdminStatus' => 'false',
      'ifNumber' => 'false',
      'ifLastChange' => 'true',
      'ifTableLastChange' => 'true',
      'skipIpAddressTableOnSingle' => 'true',
    }
  },

Model custom - interface - ifAdminStatus

Default: true

Every poll cycle NMIS is checking the ifAdminStatus and ifOperStatus to see if interfaces have changed state or not.  This is not a problem with regular nodes, but when slow SNMP agents combine with high interface counts, this can take extra time.  In the model for the node you can disable this processing by adding a top level modelling section as below.  NMIS will still check the interface status of interfaces being collected but will not look for interfaces which have changed state.

    'custom' => {
		'interface' => {
		  'ifAdminStatus' => 'false',
		}
	},

Polling Locks

...

Model custom - interface - ifNumber

Default: true

Every poll cycle the number of interfaces is verified using the SNMP MIB ifNumber, if this value has changed from the last poll, an interface has been added or deleted and by default NMIS will run an interface update to recalculate what needs to be polled.

Model custom - interface - ifLastChange

Default: false

If disabling ifAdminStatus, you really need a way to check that interfaces have changed or not, which is done with this option set to true.  This will use the mib value ifLastChange to determine if an interface has changed state and then run an update on that interface alone.

The description of this MIB variable is: "The value of sysUpTime at the time the interface entered its current operational state. If the current state was entered prior to the last re-initialization of the local network management subsystem, then this object contains a zero value."

Model custom - interface - ifTableLastChange

Default: false

When enabled, the interface update process for all interfaces will be skipped if this value has not changed.  NMIS will collect this value from the node and compare it to the last time it was collected, if there time is newer than it will run a full interface update.  

The SNMP variable ifTableLastChange holds the value of sysUpTime when the interface table last changed, that is when an interface was added or deleted.  The default value for this is false.  When set to true the NMIS update process will check this value for each node and only run a complete update of the interface table when this value has changed.  For a node with high interface counts, this can result in not needing to run an update very often at all.  Which tends to work well as these devices do not often have new interface being added.

The description of this MIB variable is: "The value of sysUpTime at the time of the last creation or deletion of an entry in the ifTable. If the number of entries has been unchanged since the last re-initialization of the local network management subsystem, then this object contains a zero value."

Model custom - interface - skipIpAddressTableOnSingle

 Default: false