eVIP Network Counters Measurements
Evolved Virtual IP

Contents

1Introduction

2
Performance Measurement Groups

3
Measurements of EvipNetstat PM Group

4
Performance Measurement Jobs

5
Measurement Job Results
5.1Example ROID or measObjLdn

1   Introduction

This document describes the PM Counters interface provided by eVIP. A number of measurements are available to report network statistics provided by eVIP.

eVIP PM counters are only available when eVIP is run in a Core MW environment and cannot be used by eVIP for LSB.

For more information about Performance Management, refer to Core MW Performance Management Description.

2   Performance Measurement Groups

eVIP has one PM Group that applies to both real and virtualized hardware.

Table 1    PM Group Names

pmGroupId

Properties

EvipNetstat

  • description: Network statistics for a particular FEE or ALB. Syntax for ROID or measObjLdn (Measured Object Local Distinguished Name) of related measurementType: Alb=<alb_name> or Alb=<alb_name>,Fee=<fee_name>. See Section 5.1

  • pmGroupVersion: 1

Note:  
Providing the FEE name is optional. When FEE is omitted, aggregated values of all FEEs within the ALB are provided.

3   Measurements of EvipNetstat PM Group

PM producer daemons (apart of evip_amf_proxy) on each eVIP node collect statistical information which is reported via the Core MW Performance Management Interface. For more details on this interface, refer to Core MW Performance Management Description. All network measurements are reported as delta (change in value) from previous sample.

Table 2    eVIP Measurement Instances

measurementTypeId

Properties

ip4OutOctets

  • description: IPv4 Octets Sent. The measured value includes data passing through FEE elements and does not include data related to routing protocols and intra-cluster traffic.

  • collectionMethod: CC

  • aggregation: SUM

ip6OutOctets

  • description: IPv6 Octets Sent. The measured value includes data passing through FEE elements and does not include data related to routing protocols and intra-cluster traffic.

  • collectionMethod: CC

  • aggregation:: SUM

ip4OutDatagrams

  • description: IPv4 Datagrams Sent. The measured value includes data passing through FEE elements and does not include data related to routing protocols and intra-cluster traffic.

  • collectionMethod: CC

  • aggregation:SUM

ip6OutDatagrams

  • description: IPv6 Datagrams Sent. The measured value includes data passing through FEE elements and does not include data related to routing protocols and intra-cluster traffic.

  • collectionMethod: CC

  • aggregation:: SUM

ip4InOctets

  • description: IPv4 Octets Received. The measured value includes data passing through FEE elements and does not include data related to routing protocols and intra-cluster traffic.

  • collectionMethod: CC

  • aggregation: SUM

ip6InOctets

  • description: IPv6 Octets Received. The measured value includes data passing through FEE elements and does not include data related to routing protocols and intra-cluster traffic.

  • collectionMethod: CC

  • aggregation: SUM

ip4InDatagrams

  • description: IPv4 Datagrams Received. The measured value includes data passing through FEE elements and does not include data related to routing protocols and intra-cluster traffic.

  • collectionMethod: CC

  • aggregation: SUM

ip6InDatagrams

  • description: IPv6 Datagrams Received. The measured value includes data passing through FEE elements and does not include data related to routing protocols and intra-cluster traffic.

  • collectionMethod: CC

  • aggregation: SUM

l2RxDropped

  • description: Number of received ethenet frames with wrong VLAN or proto id unsupported/not configured on this interface

  • collectionMethod: CC

  • aggregation: SUM

l2RxErrors

  • description: Number of received ethernet frames with size not divisible by 8

  • collectionMethod: CC

  • aggregation: SUM

l2RxFifoErrors

  • description: Number of received frames which were received when queue from device driver to IP stack in kernel was full i.e. no of ingress overruns

  • collectionMethod: CC

  • aggregation: SUM

l2RxFrameErrors

  • description: Number of frames received with CRC errors, too-short frames and too-long frames

  • collectionMethod: CC

  • aggregation: SUM

l2TxCarrierErrors

  • description: Number of carrier sense errors detected on wire

  • collectionMethod: CC

  • aggregation: SUM

l2TxCollisions

  • description: Number of times CSMA/CD collisions were detected on wire

  • collectionMethod: CC

  • aggregation: SUM

l2TxDropped

  • description: Number of frames dropped by device driver

  • collectionMethod: CC

  • aggregation: SUM

l2TxErrors

  • description: Number of ethernet frame sending errors reported by device driver

  • collectionMethod: CC

  • aggregation: SUM

l2TxFifoErrors

  • description: Number of received frames which were received when queue from device driver to IP stack in kernel was full i.e. no of ingress overruns

  • collectionMethod: CC

  • aggregation: SUM

4   Performance Measurement Jobs

Measurement jobs can be created as ECIM PM, Use Case Description. The COM CLI can be used to create, activate, modify, and delete PM jobs as per examples below.

Example 1   Example via ECIM, to set up a new measurement job:

configure
ManagedElement=1,SystemFunctions=1,Pm=1
PmJob=MyEvipJob
requestedJobState=STOPPED
granularityPeriod=ONE_MIN
reportingPeriod=ONE_MIN
MeasurementReader=1
measurementSpecification,groupRef="ManagedElement=1,SystemFunctions=1,Pm=1,PmGroup=EvipNetstat"
commit
show
exit

Example 2   Example via ECIM, to activate the measurement job:

configure
ManagedElement=1,SystemFunctions=1,Pm=1
PmJob=MyEvipJob
requestedJobState=ACTIVE
commit
show
exit

Example 3   Example via ECIM, to change the reporting and granularity period to five minutes:

configure
ManagedElement=1,SystemFunctions=1,Pm=1
PmJob=MyEvipJob
granularityPeriod=FIVE_MIN
reportingPeriod=FIVE_MIN
commit
exit

Example 4   Example via ECIM, to delete a measurement job:

configure
ManagedElement=1,SystemFunctions=1,Pm=1
PmJob=MyEvipJob
requestedJobState=STOPPED
commit
up
configure
no PmJob=MyEvipJob
commit
exit

Alternatively, Core MW command line tools can be used to create a job as per examples below.

Example 5   Example via Core MW commands:

cmw-pmjob-create MyEvipJob -u EvipNetstat -p3 -t1 -q1 -g3 -r3
cmw-pm-show-counters Alb=alb1,Fee=fee1

See further relevant command line tools provided by Core MW:
cmw-pmjob-create, cmw-pmjob-delete, cmw-pmjob-list, cmw-pmjob-modify, cmw-pmjob-start, cmw-pmjob-status,
cmw-pmjob-stop, cmw-pm-show-counters.
Note:  
  • Due to Core MW PM architecture and its method of collection, changes to FEE allocation, such as reboot, shutdown, scaling-in, can cause up to one granularity period of packet counter data to be lost.
  • When adding a job, the granularity_period should be equal to reporting_period. Periods below one minute are not possible.

5   Measurement Job Results

For measObjLdn, see Section 5.1.

Here are some examples:

5.1   Example ROID or measObjLdn

The ROID or measObjLdn (Measured Object Local Distinguished Name) can be constructed for each measurement instance by using the correct identifiers after the appropriate PM Group.

For example

Legal | © Ericsson AB 2018