Ticket 13084

Summary:	Partition/gres for two network interfaces
Product:	Slurm	Reporter:	stefano.carrazza
Component:	Configuration	Assignee:	Jacob Jenson <jacob>
Status:	RESOLVED INVALID	QA Contact:
Severity:	6 - No support contract
Priority:	---	CC:	stefano.carrazza
Version:	22.05.x
Hardware:	Linux
OS:	Linux
Site:	-Other-	Alineos Sites:	---
Atos/Eviden Sites:	---	Confidential Site:	---
Coreweave sites:	---	Cray Sites:	---
DS9 clusters:	---	HPCnow Sites:	---
HPE Sites:	---	IBM Sites:	---
NOAA SIte:	---	OCF Sites:	---
Recursion Pharma Sites:	---	SFW Sites:	---
SNIC sites:	---	Linux Distro:	---
Machine Name:		CLE Version:
Version Fixed:		Target Release:	---
DevPrio:	---	Emory-Cloud Sites:	---

Description stefano.carrazza 2021-12-27 03:04:42 MST

Dear developers,

we are considering SLURM in our setup where we would like to provide a single node with 2 network interfaces: the default network and a local custom network.

We would like to have a special partition (or gres) which accepts jobs with access to the custom local network. This partition/gres should execute just one job at time, but the machine should be able to run jobs from other partitions (connected to the default network) simultaneously.

We would like to know if SLURM provides a standard procedure for this setup.

Comment 1 Jacob Jenson 2021-12-28 12:59:14 MST

Stefano, 

Yes, there are standard procedures for this. 

If you would like SchedMD to help set this up please let me know and we can put a Slurm support contact in place. 

Jacob