FlexPod Datacenter For AI/ML With Cisco UCS 480 ML For Deep Learning

1y ago
27 Views
3 Downloads
5.14 MB
145 Pages
Last View : 19d ago
Last Download : 2m ago
Upload by : Mollie Blount
Transcription

FlexPod Datacenter for AI/ML with CiscoUCS 480 ML for Deep LearningDeployment Guide for FlexPod Datacenter for AI/ML withCisco UCS 480 ML for Deep LearningPublished: January 16, 2020

About the Cisco Validated Design ProgramThe Cisco Validated Design (CVD) program consists of systems and solutions designed, tested, and documentedto facilitate faster, more reliable, and more predictable customer deployments. For more information, go to:http://www.cisco.com/go/designzone.ALL DESIGNS, SPECIFICATIONS, STATEMENTS, INFORMATION, AND RECOMMENDATIONS (COLLECTIVELY,"DESIGNS") IN THIS MANUAL ARE PRESENTED "AS IS," WITH ALL FAULTS. CISCO AND ITS SUPPLIERSDISCLAIM ALL WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE WARRANTY OF MERCHANTABILITY,FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OR ARISING FROM A COURSE OF DEALING,USAGE, OR TRADE PRACTICE. IN NO EVENT SHALL CISCO OR ITS SUPPLIERS BE LIABLE FOR ANY INDIRECT,SPECIAL, CONSEQUENTIAL, OR INCIDENTAL DAMAGES, INCLUDING, WITHOUT LIMITATION, LOST PROFITS ORLOSS OR DAMAGE TO DATA ARISING OUT OF THE USE OR INABILITY TO USE THE DESIGNS, EVEN IF CISCOOR ITS SUPPLIERS HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.THE DESIGNS ARE SUBJECT TO CHANGE WITHOUT NOTICE. USERS ARE SOLELY RESPONSIBLE FOR THEIRAPPLICATION OF THE DESIGNS. THE DESIGNS DO NOT CONSTITUTE THE TECHNICAL OR OTHERPROFESSIONAL ADVICE OF CISCO, ITS SUPPLIERS OR PARTNERS. USERS SHOULD CONSULT THEIR OWNTECHNICAL ADVISORS BEFORE IMPLEMENTING THE DESIGNS. RESULTS MAY VARY DEPENDING ONFACTORS NOT TESTED BY CISCO.CCDE, CCENT, Cisco Eos, Cisco Lumin, Cisco Nexus, Cisco StadiumVision, Cisco TelePresence, Cisco WebEx,the Cisco logo, DCE, and Welcome to the Human Network are trademarks; Changing the Way We Work, Live,Play, and Learn and Cisco Store are service marks; and Access Registrar, Aironet, AsyncOS, Bringing the MeetingTo You, Catalyst, CCDA, CCDP, CCIE, CCIP, CCNA, CCNP, CCSP, CCVP, Cisco, the Cisco Certified InternetworkExpert logo, Cisco IOS, Cisco Press, Cisco Systems, Cisco Systems Capital, the Cisco Systems logo, CiscoUnified Computing System (Cisco UCS), Cisco UCS B-Series Blade Servers, Cisco UCS C-Series Rack Servers,Cisco UCS S-Series Storage Servers, Cisco UCS Manager, Cisco UCS Management Software, Cisco UnifiedFabric, Cisco Application Centric Infrastructure, Cisco Nexus 9000 Series, Cisco Nexus 7000 Series. Cisco PrimeData Center Network Manager, Cisco NX-OS Software, Cisco MDS Series, Cisco Unity, Collaboration WithoutLimitation, EtherFast, EtherSwitch, Event Center, Fast Step, Follow Me Browsing, FormShare, GigaDrive,HomeLink, Internet Quotient, IOS, iPhone, iQuick Study, LightStream, Linksys, MediaTone, MeetingPlace,MeetingPlace Chime Sound, MGX, Networkers, Networking Academy, Network Registrar, PCNow, PIX,PowerPanels, ProConnect, ScriptShare, SenderBase, SMARTnet, Spectrum Expert, StackWise, The Fastest Wayto Increase Your Internet Quotient, TransPath, WebEx, and the WebEx logo are registered trademarks of CiscoSystems, Inc. and/or its affiliates in the United States and certain other countries.All other trademarks mentioned in this document or website are the property of their respective owners. The useof the word partner does not imply a partnership relationship between Cisco and any other company. (0809R) 2020 Cisco Systems, Inc. All rights reserved.2

Table of ContentsExecutive Summary . 9Solution Overview . 10Introduction . 10Audience . 10What’s New in this Release? . 10Solution Design . 11Architecture . 11Physical Topology . 12Base Infrastructure .13Hardware and Software Revisions . 13Required VLANs . 14Physical Infrastructure .14FlexPod Cabling .14Network Switch Configuration . 16vGPU-only Deployment in Existing VMware Environment . 16Enable Features . 16Cisco Nexus A and Cisco Nexus B . 16Global Configurations.17Cisco Nexus A and Cisco Nexus B . 17Create VLANs . 17Cisco Nexus A and Cisco Nexus B . 17Configure Virtual Port-Channel Parameters .18Cisco Nexus A . 18Cisco Nexus B . 18Configure Virtual Port-Channels . 19Cisco UCS 6454 Fabric Interconnect to Nexus 9336C-FX2 Connectivity . 19NetApp A800 to Nexus 9336C-FX2 Connectivity . 22Storage Configuration .25Remove Ports from Default Broadcast Domain . 25Disable flow control on 100GbE ports. . 25Disable Auto-Negotiate on 100GbE Ports .26Enable Cisco Discovery Protocol . 26Enable Link-layer Discovery Protocol on all Ethernet Ports . 26Create Management Broadcast Domain .26Create NFS Broadcast Domain . 263

Create Interface Groups .27Change MTU on Interface Groups . 27Create VLANs . 27Configure Network Time Protocol . 28Configure SNMP . 28Configure SNMPv1 Access . 28Create SVM . 28Create Load-Sharing Mirrors of SVM Root Volume . 29Create iSCSI Service .29Configure HTTPS Access . 29Configure NFSv3 . 29Create ONTAP FlexGroup Volume . 30Create FlexVol Volumes for Bare-Metal Hosts . 30Create FlexVol Volumes for VMware ESXi Hosts . 30Create Bare-Metal Server Boot LUNs . 30Create ESXi Boot LUNs .31Modify Volume Efficiency . 31Schedule Deduplication .31Create NFS LIFs . 31Create iSCSI LIFs . 32Add AI-ML SVM Administrator. 32Cisco UCS Configuration for VMware with vGPU . 33Cisco UCS Base Configuration . 33Create NFS VLAN. 33ADD VLAN to vNIC Template . 33VMware Setup and Configuration for vGPU .35Obtaining and installing NVIDIA vGPU Software . 35NVIDIA Licensing .35Download NVIDIA vGPU Software . 36Setup NVIDIA vGPU Software License Server. 36Register License Server to NVIDIA Software Licensing Center .38Install NVIDIA vGPU Manager in ESXi . 41Set the Host Graphics to SharedPassthru .42(Optional) Enabling vMotion with vGPU. 43Add a Port-Group to access AI/ML NFS Share . 44Red Hat Enterprise Linux VM Setup. 45VM Hardware Setup .464

Download RHEL 7.6 DVD ISO . 47Operating System Installation . 48Network and Hostname Setup . 49RHEL VM – Base Configuration . 53Log into RHEL Host using SSH. 54Setup Subscription Manager . 54Enable Repositories .54Install Net-Tools and Verify MTU. 54Install FTP . 55Enable EPEL Repository . 55Install NFS Utilities and Mount NFS Share .55Setup NTP . 57Disable Firewall . 57Disable IPv6 (Optional) . 57Install Kernel Headers .58Install gcc . 58Install wget . 59Install DKMS . 59NVIDIA and CUDA Drivers Installation . 59Add vGPU to the VM .60Install NVIDIA Driver .61Install CUDA Toolkit .62Verify the NVIDIA and CUDA Installation .64Verify CUDA Driver .64Verify NVIDIA Driver .65Setup NVIDIA vGPU Licensing on the VM .66Cisco UCS Configuration for Bare Metal Workload . 68Cisco UCS Base Configuration . 68Cisco UCS C220 M5 Connectivity . 68Enable Server Ports .69Cisco UCS C240 M5 Connectivity . 69Enable Server Ports .70Cisco UCS C480 ML M5 Connectivity . 70Enable Server Ports .70Create an IQN Pool for iSCSI Boot . 71Create iSCSI Boot IP Address Pools . 72Create MAC Address Pools. 745

Create UUID Suffix Pool .75Create Server Pool .76Create VLANs . 77Modify Default Host Firmware Package . 79Set Jumbo Frames in Cisco UCS Fabric .80Create Local Disk Configuration Policy . 81Create Network Control Policy to Enable Link Layer Discovery Protocol (LLDP). 82Create Power Control Policy . 83Create Server BIOS Policy. 84Update the Default Maintenance Policy . 86Create vNIC Templates .87Create Management vNIC Template . 88Create iSCSI Boot vNIC Templates . 89Create NFS vNIC Template . 90(Optional) Create Traffic vNIC Template .91Create LAN Connectivity Policy for iSCSI Boot. 92Create iSCSI Boot Policies . 94Create Service Profile Template . 95Configure Storage Provisioning . 95Configure Networking Options . 96Configure SAN Connectivity Options. 96Configure Zoning Options . 97Configure vNIC/HBA Placement . 97Configure vMedia Policy . 97Configure Server Boot Order . 97Configure Maintenance Policy . 100Configure Server Assignment . 101Configure Operational Policies . 101Create Service Profiles . 101Storage Configuration – Boot LUNs . 103ONTAP Boot Storage Setup . 103Create igroups . 103Map Boot LUNs to igroups . 103Bare Metal Server Setup and Configuration . 104Red Hat Enterprise Linux (RHEL) Bare Metal Installation . 104Download RHEL 7.6 DVD ISO . 104Log into Cisco UCS Manager . 1046

Operating System Installation

Cisco UCS S-Series Storage Servers, Cisco UCS Manager, Cisco UCS Management Software, Cisco Unified Fabric, Cisco Application Centric Infrastructure, Cisco Nexus 9000 Series, Cisco Nexus 7000 Series. Cisco Prime Data Center Network Manager, Cisco NX-OS Software, Cisco MDS Series, Cisco Unity, Collaboration Without

Related Documents:

Deviations based on customer requirements from a given CVD or NVA are permitted if these variations do not create an unsupported configuration. The FlexPod program includes two solutions: FlexPod Express and FlexPod Datacenter. FlexPod Express. Offers customers an entry-level solution with technologies from

Flexpod: Cisco NetApp was ist ein Flexpod? Flexpod ist eine gemeinsame Architektur (CVD / NVA) Flexpod ist lösungsortientiert Die Flexpod Architektur ist vollständig dokumentiert: Design Guide Sizing Guide Wiring diagrams Bill of Materials Deployment Guide und, und, und

FlexPod Datacenter with VMware vSphere 6.5, NetApp AFF A-Series and Fibre Channel Deployment Guide for FlexPod Datacenter with Fibre Chan-nel SAN and VMware vSphere 6.5 and ONTAP 9.1 Last Updated: August 30, 2018

Bruksanvisning för bilstereo . Bruksanvisning for bilstereo . Instrukcja obsługi samochodowego odtwarzacza stereo . Operating Instructions for Car Stereo . 610-104 . SV . Bruksanvisning i original

FlexPod configuration is racked, powered, and cabled per the instructions in the FlexPod Datacenter with Fibre Channel Storage using VMware vSphere 6.5 Update 1, NetApp AFF A-series and Cisco UCS Manager 3.2 CVD. Cisco Nexus switch configuration A fault- tolerant pair of Cisco Nexus 9300

Each Cisco Validated Design (CVD) or NetApp Verified Architecture (NVA) is a possible FlexPod configuration. Cisco and NetApp document these configuration combinations and validate them with extensive end-to-end testing. The FlexPod dep

FLEXPOD ERFOLGREICH 10 JAHRE FLEXPOD 100 CVDS ABER 700 MANUELLE TASKS 4 ELEMENT MANAGER HIGH SKILLED ENGINEERS WERDEN BENÖTIGT CVD: FlexPod Datacenter with

(WANs) [24, 26, 27]. WAN traffic shares the datacenter network with intra-datacenter traffic, with the ratio of datacenter to WAN traffic typically around 5:1 [42]. Despite the small fraction of WAN traffic, we find that its impact on datacenter traffic is significant when both ty