Difference between revisions of "Deduplication in GigaFlow"

From Observer GigaFlow Support | VIAVI Solutions Inc.
Jump to: navigation, search
(= Changing the Default Settings)
(= Changing the Default Settings)
Line 3: Line 3:
 
GigaFlow can deduplicate flows although it is not recommended by default. Duplicate flows can be used for forensic analysis, e.g. in the situation where an IP pair is received by a router on an ingress port, deduplication reduces the information available for trouble-shooting.
 
GigaFlow can deduplicate flows although it is not recommended by default. Duplicate flows can be used for forensic analysis, e.g. in the situation where an IP pair is received by a router on an ingress port, deduplication reduces the information available for trouble-shooting.
  
=== Changing the Default Settings ==
+
== Changing the Default Settings ==
  
 
GigaFlow's deduplication behaviour can be modified at '''System > Global > General'''.
 
GigaFlow's deduplication behaviour can be modified at '''System > Global > General'''.

Revision as of 11:56, 27 September 2019

Overview

GigaFlow can deduplicate flows although it is not recommended by default. Duplicate flows can be used for forensic analysis, e.g. in the situation where an IP pair is received by a router on an ingress port, deduplication reduces the information available for trouble-shooting.

Changing the Default Settings

GigaFlow's deduplication behaviour can be modified at System > Global > General.

  • Setting Process Duplicate Flows to No: will disable all processing of duplicate flows, i.e. no storage, no forensics, no summaries. These will be accounted for as increments to the duplicates counter on the device.
  • Setting Process Duplicate Flows to Yes and Ignore Duplicates on Summaries to Yes: will fully process and store the duplicates but the byte/flow/packet count against interfaces will not be affected by the duplicates. This is important for capacity planning.
  • Setting Process Duplicate Flows to Yes and Ignore Duplicates on Summaries to No: will fully process and store the duplicates and the byte/flow/packet count against interfaces will be affected by the duplicates. This is not good for capacity planning.