What is Halo nodes in SU2?

davidsubo · October 23, 2018, 19:04

Hi. I start to learn SU2_PY. For FSI problems,in FSIinterface.py, there is line:

self.nLocalFluidInterfacePhysicalNodes = self.nLocalFluidInterfaceNodes - self.nLocalFluidInterfaceHaloNode

I am confused what is HaloNode of fluid mesh? is it connected with dual mesh?

Does anybody know this and is there figure show this? I am looking forward to your response.

Thank you very much!

pcg · October 24, 2018, 11:02

Halo nodes are created during partitioning of the domain for parallelization.
Each MPI rank (i.e. processor) gets a part of the original domain. These smaller parts overlap slightly, some nodes appear in two partitions, one "owns" the node the other does not, for the latter that node is a halo node.

davidsubo · October 25, 2018, 00:02

Thank you! Yes,it is for information transfer between the the partitions for parallelization.

October 23, 2018, 19:04	What is Halo nodes in SU2?	#1
davidsubo New Member bosu Join Date: Aug 2018 Posts: 2 Rep Power: 0	Hi. I start to learn SU2_PY. For FSI problems,in FSIinterface.py, there is line: self.nLocalFluidInterfacePhysicalNodes = self.nLocalFluidInterfaceNodes - self.nLocalFluidInterfaceHaloNode I am confused what is HaloNode of fluid mesh? is it connected with dual mesh? Does anybody know this and is there figure show this? I am looking forward to your response. Thank you very much! Last edited by davidsubo; October 23, 2018 at 23:09.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to run SU2 on cluster of computers (How to specify nodes ?)	aero_amit	SU2	29	February 24, 2020 14:44
Introducing SU2 International Developers Society (IDS)	fpalacios	SU2 News & Announcements	1	June 17, 2019 23:38
SU2 code scaling poorly on multiple nodes	Samirs	SU2	1	August 25, 2018 20:15
Welcome to the Stanford University Unstructured (SU2) forum!	economon	SU2	0	January 7, 2013 03:48
CFX4.3 -build analysis form	Chie Min	CFX	5	July 13, 2001 00:19

October 24, 2018, 11:02		#2
pcg Senior Member Pedro Gomes Join Date: Dec 2017 Posts: 466 Rep Power: 13	Halo nodes are created during partitioning of the domain for parallelization. Each MPI rank (i.e. processor) gets a part of the original domain. These smaller parts overlap slightly, some nodes appear in two partitions, one "owns" the node the other does not, for the latter that node is a halo node.

October 25, 2018, 00:02		#3
davidsubo New Member bosu Join Date: Aug 2018 Posts: 2 Rep Power: 0	Thank you! Yes,it is for information transfer between the the partitions for parallelization.