A importância do Sizing no sucesso do HCI...Management Tools 2 1 Innovative 3 Centralized...
Transcript of A importância do Sizing no sucesso do HCI...Management Tools 2 1 Innovative 3 Centralized...
A importância do Sizing no sucesso do HCI
Amauri Barros @amauripbSystem Engineer [email protected]@gmail.com+55 19 99833 9784
as informações aqui contidas são pessoais
e não representam o meu empregador e não
necessariamente representam a opinião da
empresa onde eu trabalho
Amauri Pereira de Barros
System Engineer Lenovo
Linkedin: https://www.linkedin.com/in/amaurib/
Twitter: @amauripb
Email corporativo: [email protected]
Email pessoal: [email protected]
Celular: +55 19 99833 9784
Agenda
Apresentação Lenovo
O que não esquecer no projeto
Performance ambiente REAL
Quem é a Lenovo?
https://www.gartner.com/en/newsroom/press-releases/2020-05-20-gartner-announces-rankings-of-the-2020-supply-chain-top-25
Industry Leadership in Security
• Strategic Business Priority– Lenovo code is digitally signed and stored in
North Carolina, US
– Member of the global Forum for Incident Response and Security Teams (FIRST)
• Designed in from the Ground up– Lenovo Trusted Platform Module (TPM) ensures only
digitally signed and authorized code is loaded
– NIST and FIPS 140-2 compliant encryption
• Accountability Across the Supply Chain – Lenovo maintains Trusted Supplier List (TSL) with
quarterly assessments
– Option to specify local manufacture
– Awarded industry’s highest security level by the US Customs & Border Protection
https://www.lenovo.com/us/en/data-center/solutions/sap
NOV 14 NOV 15 NOV 16 NOV 17 NOV 18
Lenovo - Provedor #1 de Supercomputadores no Mundo
Em 20 países
Posição
0
180
sistemas
JUN 18
36%
participação
#1 emperformance
agregada
JUN 19
NOV 19
JUN 20https://www.top500.org/statistics/list/ - June/2020
Confiabilidade dos Servidores Lenovo
LÍDER EM
RESILIÊNCIA
Até
Mais economia por ano
devido a alta
disponibilidade dos
servidores Lenovo
quando comparado com
outras plataformas
34x
#1 X86 POR 12 ANOS
https://lenovopress.com/lp1117-itic-reliability-study
Veeam + Lenovo
Addressing business continuity and resiliency challenges
https://go.veeam.com/veeamon-tour-2020-latam-br.html
U$250
Partner of the Year Award
Value
2020 Global Winner
https://www.lenovoandvmware.com/
The Lenovo solution for VMware SDDC
provides businesses with an affordable,
interoperable, and reliable industry-leading
cloud solution to manage all of their
virtualized workloads.
Built around the latest Lenovo ThinkAgile
VX certified nodes and appliances
https://lenovopress.com/lp0661-reference-architecture-vmware-software-defined-data-center-thinkagile-vx
• VX 7520 | 2Us
• VX 4Us HANA
VX Product Options
• VX3320 | 1U
• VX5520 | 2Us
https://lenovopress.com/lp1136-thinkagile-vx3320-appliance-xeon-sp-gen2
https://lenovopress.com/lp1142-thinkagile-vx-1u-certified-node-xeon-sp-gen2
https://lenovopress.com/lp1139-thinkagile-vx5520-appliance-xeon-sp-gen2
https://lenovopress.com/lp1143-thinkagile-vx-2u-certified-node-xeon-sp-gen2
https://lenovopress.com/lp1141-thinkagile-vx7520-appliance-xeon-sp-gen2
https://lenovopress.com/lp1143-thinkagile-vx-2u-certified-node-xeon-sp-gen2
https://lenovopress.com/lp1341-thinkagile-vx-4u-certified-node-sap-hana-gen2
https://www.lenovo.com/br/pt/data-center/services/TruScale-Infrastructure-Services/p/truscale-infrastructure-services
Sizing do Projeto
3 camadas para HCI
Servidor - Proc +
Memoria
SW Virtualização de
Servidores
Rede + SW de
Gerenciamento
Armazenamento
Controladora
+Discos +SW de
Gerenciamento
Servidor - Proc +
Memoria
+Armazenamento
SW Virtualização
+SW Armazenamento
Rede + SW de
Gerenciamento
Ready!
Quantidade de nós
https://storagehub.vmware.com/t/vmware-r-vsan-tm-design-and-sizing-guide-2/
https://storagehub.vmware.com/t/vsan-space-efficiency-technologies/host-requirements-1/
2-Node vSAN ROBO
• VMware HCI Kit ROBO (per-25 VMs)– ~x2 à 3x
• VMware HCI Kit (per-CPU)
https://cormachogan.com/2017/10/06/2-node-vsan-witness-network-design-considerations/
https://storagehub.vmware.com/t/vsan-2-node-guide/vsan-witness-appliance-sizing/
Processadores – x1xx, x2xx, agora x3xx
https://www.intel.com/content/www/us/en/products/docs/processors/xeon/3rd-gen-xeon-scalable-processors-brief.html
https://ark.intel.com/content/www/us/en/ark/products/series/204098/3rd-generation-intel-xeon-scalable-processors.html
https://xeonprocessoradvisor.intel.com/exodus/login
Processadores – Consideraçõs
• 10% CPU overhead para vSAN/vSphere
• Muitas vezes é melhor cenário de:– 4x hosts de 1-CPU
VS
– 3x hosts de 2-CPUs
• O que vcs usam de pCORE:vCPU?– 1:6, 1:10?
– É comum SAP HANA, VoliP Unify/Cisco/Avaya pedirem 1:1
• Importante pensar em falha de um nó, picos, crescimento futuro
https://docs.vmware.com/en/VMware-vSphere/7.0/com.vmware.vsphere.vsan-
planning.doc/GUID-07EFD36A-F844-4E7D-830D-3863E4AA617C.html
Processadores vs NVMe
• Second processor enables the onboard NVMe controller
https://lenovopress.com/lp1050-thinksystem-sr650-server-xeon-sp-gen2
Memória
• SEMPRE mais é melhor, NO oversubscription
• Considerar através de alguma ferramenta o overhead
• Considerar, se possível, o máximo dos canais de comunicação
• Pentes de 256, 128 – Ainda não é realidade
• Pentes de 64, 32 – É do dia-a-dia
• Pentes de 16, 8 – NO PLEASE
https://kb.vmware.com/s/article/2113954
https://vsansizer.vmware.com/
Não se esqueça de outros overheads
• vCenter / vSphere Replication
• Backup Server e/ou Proxies
• vRealize Suite Lifecycle Manager
• vRealize Operations / LogInsight
• vRealize Network Insight
• vRealize Automation / Identity Manager
• Managers e Edges do NSX-T
• SDDC Manager (VCF)
Ainda sobre… VCF:Cluster External Services
https://docs.vmware.com/en/VMware-Cloud-Foundation/3.9/com.vmware.vcf.planprep.doc_39/GUID-F022BD3C-F11C-4EE6-83EA-ABE016E7A9B9.html
Armazenamento – HBA
• Controladorassss
https://storagehub.vmware.com/t/vmware-r-vsan-tm-design-and-sizing-guide-2/choosing-a-storage-i-o-controller-1/
Armazenamento – Discos
• NVMe – espetacular
• SSD-SAS, ótimo
• SSD-SATA, muito bom
• HDD10k-SAS, bom, pouco usado
• HDD7,2K-NL-SAS, com cache adequado vai muito bem
• Não é suportado discos 7,2K SATA
• Mix de cache NVMe e capacidade SSD-SATA resolve a maioria dos casos
• NL-SAS bem dimensionado já substituiu muito v7000, VNX, Unity, FAS, 3Par…
Armazenamento – Cache
• Reads– In both hybrid and all-flash configurations vSAN checks to see if the requested block is still
hot in the cache tier. If so, this is called a cache hit. vSAN handles cache misses differently for hybrid and all-flash.
– As mentioned, in a hybrid configuration if the block is present, the read is serviced from the read cache. If a read miss occurs, vSAN will retrieve the data from the capacity tier and serve it up to the requesting application. vSAN also has a read ahead cache optimization where 1 MB of data around the data block being read is also brought into the cache. The assumption here is that next read will likely be local to the last read and will now also be cached.
– In an all-flash configuration, there is no read cache. If a requested block is in the write buffer, the request will be served from there. If not, vSAN will read the data from the capacity tier. Since the capacity tier is all-flash the impact is minimal. By not implementing a read cache on all-flash configurations the cache tier can handle more writes, boosting overall performance.
• Writes– In both hybrid and all-flash configurations, the write cache acts as a write-back buffer. When
an application issues a write operation, the write is sent to appropriate ESXi host cache based on the storage policy (i.e. Failures to tolerate, stripes, RAID, etc).
– In a hybrid configuration, 30% of the cache tier is dedicated to write buffering. Writes in the buffer are acknowledged back to the VM without having to be moved to the capacity tier first.
– In an all-flash configuration, 100% of the cache device is dedicated to write-buffering (up to a maximum of 600 GB). vSAN still utilizes the entire disk regardless of size spreading the writes to every block on the device. This reduces the wear of the cells on the flash device, ultimately increasing the life-span.
https://blogs.vmware.com/virtualblocks/2019/04/18/vsan-disk-groups/
10% vSAN rule caching, calculate on VM basis not disk capacity!
http://www.yellow-bricks.com/2016/02/16/10-rule-vsan-caching-calculate-vm-basis-not-disk-capacity/
• CACHE:
• 12x 800GB SAS SSD
• 9.6TB Cache
• +10% Recomendação
• DADOS:
• 24x 8TB NL-SAS
Armazenamento – Capacidade
• SLACK – de 20% à 30%
• Se não tiver espaço livre, o processo
de “rolling update” não vai
ISSO É CRITICO
• Cuidado no gerenciamento de
capacidade, bruto vs líquido
• É o ponto de maior atenção tanto antes
como após o projeto
https://kauteetech.github.io/vsancapacity/
Armazenamento – Performance
150 VMs *
350 IOPS
= 52.500
vsansizer.vmware.com
Networking – Hardware
• Apesar de documentado 1Gbps para Hibrido, nunca vi, alguém já?
• Dual 10Gbps, Quad 10Gbps… 25Gbps já está aí
• https://www.mellanox.com/files/doc-2020/br-sn2000-series.pdf
• “Supports flat latency of 300ns in cut-through mode”
https://blogs.vmware.com/virtualblocks/2019/04/21/designing-vsan-networks-2019-update/
https://blogs.vmware.com/virtualblocks/2018/02/28/reliable-network-connectivity-hyper-converged-environments/
vSphere Enterprise Plus vs Standard
Distributed
Resource
Scheduler
vSphere Enterprise Plus vs Standard
• Com qualquer edição do vSAN o vSphere “ganha” o Switch Distribuido
https://www.vmware.com/content/dam/digitalmarketing/vmware/en/pdf/products/vsan/vmware-vsan-licensing-guide.pdf
vsansizer.vmware.com
Premium Subscription Promotion
https://mylearn.vmware.com/descriptions/VLZ-Premium-Subscription-6-Month-Promotion-External-FAQ.pdf
https://blogs.vmware.com/education/files/2020/05/VLZ-Premium-Subscription-6-Month-Promotion-External-FAQ-updated-05132020.pdf
Embedded Platform
Management
▪ Embedded management engine common in all ThinkSystem and ThinkAgile
▪ Fresh, uncluttered graphical user interface▪ Redfish-compliant web-based REST APIs
for ease of inter-operability
Platform
Management Tools 2
1
3Innovative
Centralized
Management
▪ Centralized software-based delivery and management for ThinkSystem and ThinkAgile,storage, and networking
▪ Mobile app for anywhere management▪ REST APIs for ease of integration into
software-defined environments
4Cloud automation
and IT service
management
processes
▪ Integration into leading virtualization management consoles and IT service management tools
▪ VMware, Microsoft, Cloudforms, Chef, Puppet, ServiceNow, MSFT WAC
▪ Collection of one to one management tools▪ Scripting tools used by large companies (
ie Morgan Stanely, SAP etc..)
Orchestrator
XClarity Family of Software
XClarity Integrator for VMware vCenter
• Consolidate virtual and physical infrastructure management using your familiar console
– Discovery
– Monitoring
– Firmware updates
– Configurations
• Eliminate downtime in vSphere clusters
– Automate rolling reboots & firmware updates
– Automate evacuation of VMs from impacted hosts based on user-defined events
.
Physical Host Physical Host Physical Host
VMware vSphere
V
M
V
M
V
M
V
M
V
M
V
M
VMware vCenter
Download XClarity Integrator for VMware vCenter
vCenter
Virtualization Management
XClarity Integrator for VMware vRA and vRO
• Abstract infrastructure resources and
transform them into services
• Deliver infrastructure faster through
repeatable, scalable execution of
tasks across software and hardware
domains
• Create and manage pools of
resources, such as provisioning end-
to-end hosts with XClarity Blueprints
vRealize Automation
Service Blueprints
vRealize Orchestrator
Workflows
Download XClarity Integrator for VMware vRealize Automation
Performance ambiente
REAL
Desenho macro do ambiente
Backup &
Replication
Veeam
repository
vSphere vSpherevSphere
SR950|Platinum|3TB|15TiB
SR950|Platinum|3TB|30TiB VX7520|Gold|1,5TB|17TiB
SR650|Silver|64GB|72TiB
NE1032CE0128
Fotos do ambiente real
Fotos do ambiente real
HCIBench Report - Configuration
• Easy Run: true
• Easy Run Workloads: 4k 70r / 30w
• Storage Policy: Default Policy
• Number of Guest VMs: 6
• Number of vmdk per VM: 8
• Size of Data Disk in GB: 14
• 14GB * 8vmdk * 6VMs = 672GBs
• vSAN Configurations– vSAN Datastore Name: vsanDatastore
– vSAN Type: All-Flash
– Number of Hosts: 3
– Disk Groups per Host: 1
– Capacity Disk per Disk Group: 5
– Deduplication/Compression Enabled: 0
– Host Primary Fault Tolerance: 1
– Host Secondary Fault Tolerance: 0
– Checksum Disabled: false
HCIBench Report - Results
Amauri Barros @amauripb
System Engineer Lenovo
+55 19 99833 9784