Release 12.0 R7 | December 2014 | 3HE 08833 AAAD TQZZA Edition 01
The event log stores BGP events and DC events. BGP events include reachability
alarms and flapping occurrences for the BGP prefix. When you perform event
retrieval on a dVRS service, you can troubleshoot historical events by correlating DC
events to BGP events. For example, if there is an Prefix Down BGP event, this means
the virtual object was likely deleted and there will be a VM Down event for the
related virtual object. In this case, an alarm would not be raised as long as the relation
between events is evident. Flapping events on the VM can occur for legitimate
reasons, which can be distinguished by the 5620 SAM. Alarms regarding VM state
changes should only be generated when needed.
The event log provides access to two correlation features for BGP and DC events:
•
root cause analysis
•
impact analysis
See the 5650 CPAM User Guide for more information on root cause analysis and
impact analysis.
DC network events monitored in the 5620 SAM include the following:
•
BGP: Next Hop Change
•
BGP: Prefix Up
•
BGP: Prefix Down
•
BGP: Prefix Flapped
•
DC: Vport Up
•
DC: Vport Down
•
DC: Vport Move
•
DC: VM Up
•
DC: VM Down
Learning correlation algorithm
The 5620 SAM uses an algorithm to determine the window in which to find a
correlated event, called a learning correlation algorithm. The algorithm dynamically
learns the correlation interval and calculates the average and standard deviations.
The algorithm takes into account variables such as time delays and 5620 SAM
release load. The learning correlation algorithm improves event and alarm
correlation reliability over that of a basic correlation algorithm.
The algorithm can be set to use preset values as opposed to dynamically learned
values.
Historical Event Correlation Manager
You can access the Historical Event Correlation Manager from the Data Center
Manager. It allows you to configure the window of time in which events can be
correlated. You can also use the Correlation Partition Manager to perform the
following:
•
configure historical retention behavior
•
configure the maximum database size and length of a historical interval
•
manually clear the historical events database
Alcatel-Lucent 5620 Service Aware Manager
5620 SAM
Nuage Virtualization User Guide
11 — Services assurance and troubleshooting
•
DC: VM Move
•
DC: Virtual Switch Up
•
DC: Virtual Switch Down
•
DC: VSwitch Controller Up
•
DC: VSwitch Controller Down
•
IGP: Upstream Connection Up
•
IGP: Upstream Connection Down
•
IGP: Upstream Redundancy
Change
11-3