SlideShare a Scribd company logo
© Heraflux Technologies® 2
@kleegeek
davidklee.net
heraflux.com
linkedin.com/in/davidaklee
Specialties / Focus Areas / Passions:
• Performance Tuning & Troubleshooting
• Virtualization
• Cloud Enablement
• Infrastructure Architecture
• Health & Efficiency
• Capacity Management Founder & Chief Architect
 What is Virtualization?
 Resources and queues
 Physical servers
 Storage and interconnects
 VM architecture
 Inside the VM
© Heraflux Technologies® 3
 Added layer
 Physical resources
 Resource queues
 Very small queue delays
 When done right…
 Shared everything architecture
© Heraflux Technologies® 4
150 GHz
CPU 4 TB
Memory
4x10GbE
Network
20 TB
Tier 1
Storage
40 TB
Tier 2
Storage
VM
16 vCPU
128 GB vRAM
VM
8 vCPU
64 GB vRAM
VM
2 vCPU
16 GB
vRAM
VM
2 vCPU
16 GB
vRAM
VM
2 vCPU
16 GB
vRAM
VM
2 vCPU
16 GB
vRAM
VM
2 vCPU
16 GB
vRAM
VM
2 vCPU
16 GB
vRAM
V I R T U A L I Z A T I O N
© Heraflux Technologies® 5
Hypervisor
CPU Scheduler
CPU
Execution
CPU Scheduling Queue
Memory Allocator
Mem
R / W
Mem Allocation Queue
Disk Scheduler
Disk
R / W
Disk Scheduling Queue
Network Scheduler
Network
Tran / Rec
Network Scheduling Queue
VM TASK
VM TASK
VM TASK
VM TASK
VM TASK
© Heraflux Technologies® 6
• Finite!
• Single compute node hardware
• Total cluster compute capacity
• Storage speed (IOPs, throughput)
• VM maximums
• Interconnect path speed
Hard Limits (Resources) Soft Limits (Queues)
• Measurable but not finite
• Memory oversubscription
• CPU scheduler contention
• Shared resource utilization
• “Noisy Neighbors”
© Heraflux Technologies® 7• 7
• 7
© Heraflux Technologies® 8
Networking
Interconnects
Storage
Physical Server
Virtualization
Operating System
SQL Server Instance
SQL Server DB
Application
© Heraflux Technologies® 9
 Most shared
 Most critical
 Most complex
 Most problematic
 Slowest piece of the stack
 Many individual points of contention
© Heraflux Technologies® 10
© Heraflux Technologies® 11
Controller
Controller
LUN
LUN
LUN
LUN
Disk Pool
T1
T2
T3
 Test raw performance
 SQLIO
Deprecated!
 DiskSpd Batch
heraflux.com/go/diskspd
 Collect metrics:
 I/Os per second (IOPs)
 Latency (ms)
 Throughput (MB/s) 0.00
10000.00
20000.00
30000.00
40000.00
50000.00
60000.00
70000.00
1 2 4 8 16 32 64 128
IOps Thread Intensity
IOps Per Operations per Thread
Sequential Read
Random Read
Sequential Write
Random Write
© Heraflux Technologies® 12
© Heraflux Technologies® 13
14
(Img src: http://www.asacomputers.com/2U-Server-Sandy-Bridge.html)
 Manufacturer
 Intel
 AMD
 Sockets
 Cores
 Speed vs. GHz
 Logical vs. Physical
 Hyperthreading
© Heraflux Technologies® 15
© Heraflux Technologies® 16
(Img src: http://frankdenneman.nl/2011/01/05/amd-magny-cours-and-esx/)
CPU CPU CPU CPU
Memory
© Heraflux Technologies® 17
(Img src: http://www.learnyourtech.com/hardware/)
 Sample Cisco UCS
memory config
 Slot & chip
placement
© Heraflux Technologies® 18
(Src: http://www.cisco.com/c/dam/en/us/products/collateral/servers-unified-computing/ucs-b-series-blade-servers/b200m4-specsheet.pdf page 44)
© Heraflux Technologies® 19
 Resource limits are easy to detect / work around
 Queue contention much harder
 Time in queue = time lost from VM
 Silent performance killer
 Everything in a VM must be scheduled
 … including idle resources
 Queue processing is not always FIFO
© Heraflux Technologies® 20
 Get physical machine configuration
 Try to fit VM inside one NUMA node
 Otherwise, balance across number
of NUMA nodes
 Test configurations for
best results
© Heraflux Technologies® 21
 Example: 16 vCPU VM
 What’s better?
 2 vSocket x 8 vCore?
 4 vSocket x 4 vCore?
 8 vSocket x 2 vCore?
 Varies by workload,
hardware
 Test it for yourself!
© Heraflux Technologies® 22
0
100000
200000
300000
400000
500000
600000
700000
800000
900000
8 16 64 256
Transactions/min Concurrent HammerDB Users
vNUMA SQL Server Scalability - 16 vCPUs - HammerDB
4socket x 4CPU 8socket x 2CPU 2socket x 8CPU
 1 vCPU to 1 pCPU is
a poor recommendation
 Queues
 Ready Time
 Co-Stop
 Ex: Client environment
 74 running VMs
 VMs @ 2-8 vCPU
 72 total pCores
 Fantastic performance
© Heraflux Technologies® 23
© Heraflux Technologies® 24
© Heraflux Technologies® 25
© Heraflux Technologies® 26
 “Right” amount of vCPU and vRAM resources
 Physical world = Size for end of life
 Virtual world = Size for right now
 Idle vCPUs will slow application VMs performance
 Repeat “right-sizing” analysis routinely
© Heraflux Technologies® 27
 Virtual disks whenever possible
 Multiple SCSI controllers
 VMware – PVSCSI
 Spread out the workload
© Heraflux Technologies® 28
 C: - Operating System
 D: - SQL Server Instance Home
 E: - System Databases (master, model, msdb) *
 F: - User Database Data (1 of X)…
 G: - User Database Log (1 of Y)…
 H: - TempDB…
 Y: - Windows Page file **
 Z: - Backups Adjust as necessary
(but stay standardized)
© Heraflux Technologies® 29
 64KB NTFS block sizes
 Set power settings to “High Performance” (CPU-Z)
 Set antivirus exclusions for SQL Server (tinyurl.com/sqlav)
 Ongoing OS-level performance metric collection
 No greater than five minute interval
 Windows Perfmon, Microsoft SCOM,
or other third-party utility
 heraflux.com/go/perfmon
© Heraflux Technologies® 30
 Goal: Maximize performance while reducing resource scheduling
 Parallelizable workloads
 Determine how parallel the workload is
 Set MaxDOP = vNUMA node core count (?)
 Cost threshold for parallelism = Not default
 Jonathan Kehayias – Tuning CTOP from Plan Cache
 bit.ly/1rTs9UX
© Heraflux Technologies® 31
 Spread out the I/O
 File groups, data files, partitions
 Parallelism with multiple active storage paths
 Reduce I/O
 Table / index compression vs. SAN compression
 In-memory constructs
 More RAM
 SSD read / write caching
 Faster I/O – All flash SAN
 And then clean up bad schemas & queries!
© Heraflux Technologies® 32
 Virtualization works.
 Equivalent performance if done right
 Efficiency in data handling
© Heraflux Technologies® 33
© Heraflux Technologies® 34
@kleegeek
davidklee.net
heraflux.com
linkedin.com/in/davidaklee
Virtual SQL Servers. Actual Performance.

More Related Content

What's hot

RanjiSample
RanjiSampleRanjiSample
RanjiSampleranjth
 
Ceph Day KL - Ceph Tiering with High Performance Archiecture
Ceph Day KL - Ceph Tiering with High Performance ArchiectureCeph Day KL - Ceph Tiering with High Performance Archiecture
Ceph Day KL - Ceph Tiering with High Performance Archiecture
Ceph Community
 
XPDDS18: Xenwatch Multithreading - Dongli Zhang, Oracle
XPDDS18: Xenwatch Multithreading - Dongli Zhang, OracleXPDDS18: Xenwatch Multithreading - Dongli Zhang, Oracle
XPDDS18: Xenwatch Multithreading - Dongli Zhang, Oracle
The Linux Foundation
 
Enabling POWER 8 advanced features on Linux
Enabling POWER 8 advanced features on LinuxEnabling POWER 8 advanced features on Linux
Enabling POWER 8 advanced features on Linux
Sebastien Chabrolles
 
XPDDS18: The Art of Virtualizing Cache Maintenance - Julien Grall, Arm
XPDDS18: The Art of Virtualizing Cache Maintenance - Julien Grall, ArmXPDDS18: The Art of Virtualizing Cache Maintenance - Julien Grall, Arm
XPDDS18: The Art of Virtualizing Cache Maintenance - Julien Grall, Arm
The Linux Foundation
 
XPDDS18: Memory Overcommitment in XEN - Huang Zhichao, Huawei
XPDDS18: Memory Overcommitment in XEN - Huang Zhichao, HuaweiXPDDS18: Memory Overcommitment in XEN - Huang Zhichao, Huawei
XPDDS18: Memory Overcommitment in XEN - Huang Zhichao, Huawei
The Linux Foundation
 
Performance: Observe and Tune
Performance: Observe and TunePerformance: Observe and Tune
Performance: Observe and Tune
Paul V. Novarese
 
TDS-16489U-R2 0215 EN
TDS-16489U-R2 0215 ENTDS-16489U-R2 0215 EN
TDS-16489U-R2 0215 EN
QNAP Systems, Inc.
 
Keith Paskett - Postgres on ZFS @ Postgres Open
Keith Paskett - Postgres on ZFS @ Postgres OpenKeith Paskett - Postgres on ZFS @ Postgres Open
Keith Paskett - Postgres on ZFS @ Postgres OpenPostgresOpen
 
Ibm aix technical deep dive workshop advanced administration and problem dete...
Ibm aix technical deep dive workshop advanced administration and problem dete...Ibm aix technical deep dive workshop advanced administration and problem dete...
Ibm aix technical deep dive workshop advanced administration and problem dete...
solarisyougood
 
XPDS16: Xen Live Patching - Updating Xen Without Rebooting - Konrad Wilk, Ora...
XPDS16: Xen Live Patching - Updating Xen Without Rebooting - Konrad Wilk, Ora...XPDS16: Xen Live Patching - Updating Xen Without Rebooting - Konrad Wilk, Ora...
XPDS16: Xen Live Patching - Updating Xen Without Rebooting - Konrad Wilk, Ora...
The Linux Foundation
 
XPDDS17: Xen-lite for ARM: Adapting Xen for a Samsung Exynos MicroServer with...
XPDDS17: Xen-lite for ARM: Adapting Xen for a Samsung Exynos MicroServer with...XPDDS17: Xen-lite for ARM: Adapting Xen for a Samsung Exynos MicroServer with...
XPDDS17: Xen-lite for ARM: Adapting Xen for a Samsung Exynos MicroServer with...
The Linux Foundation
 
제3회난공불락 오픈소스 인프라세미나 - lustre
제3회난공불락 오픈소스 인프라세미나 - lustre제3회난공불락 오픈소스 인프라세미나 - lustre
제3회난공불락 오픈소스 인프라세미나 - lustre
Tommy Lee
 
XPDDS17: How to Abstract Hardware Acceleration Device in Cloud Environment - ...
XPDDS17: How to Abstract Hardware Acceleration Device in Cloud Environment - ...XPDDS17: How to Abstract Hardware Acceleration Device in Cloud Environment - ...
XPDDS17: How to Abstract Hardware Acceleration Device in Cloud Environment - ...
The Linux Foundation
 
Design decision nfs-versus_fc_storage v_0.3
Design decision nfs-versus_fc_storage v_0.3Design decision nfs-versus_fc_storage v_0.3
Design decision nfs-versus_fc_storage v_0.3David Pasek
 
Relax and Recover on POWER (Updated 05-2017)
Relax and Recover on POWER (Updated 05-2017)Relax and Recover on POWER (Updated 05-2017)
Relax and Recover on POWER (Updated 05-2017)
Sebastien Chabrolles
 
KVM Tuning @ eBay
KVM Tuning @ eBayKVM Tuning @ eBay
KVM Tuning @ eBay
Xu Jiang
 
IMCSummit 2015 - Day 2 Developer Track - The NVM Revolution
IMCSummit 2015 - Day 2 Developer Track - The NVM RevolutionIMCSummit 2015 - Day 2 Developer Track - The NVM Revolution
IMCSummit 2015 - Day 2 Developer Track - The NVM Revolution
In-Memory Computing Summit
 
XPDDS18: NVDIMM Overview - George Dunlap, Citrix
XPDDS18: NVDIMM Overview - George Dunlap, Citrix XPDDS18: NVDIMM Overview - George Dunlap, Citrix
XPDDS18: NVDIMM Overview - George Dunlap, Citrix
The Linux Foundation
 
XPDS13: Perf Support in Xen - Boris Ostrovsky, Oracle
XPDS13: Perf Support in Xen - Boris Ostrovsky, OracleXPDS13: Perf Support in Xen - Boris Ostrovsky, Oracle
XPDS13: Perf Support in Xen - Boris Ostrovsky, Oracle
The Linux Foundation
 

What's hot (20)

RanjiSample
RanjiSampleRanjiSample
RanjiSample
 
Ceph Day KL - Ceph Tiering with High Performance Archiecture
Ceph Day KL - Ceph Tiering with High Performance ArchiectureCeph Day KL - Ceph Tiering with High Performance Archiecture
Ceph Day KL - Ceph Tiering with High Performance Archiecture
 
XPDDS18: Xenwatch Multithreading - Dongli Zhang, Oracle
XPDDS18: Xenwatch Multithreading - Dongli Zhang, OracleXPDDS18: Xenwatch Multithreading - Dongli Zhang, Oracle
XPDDS18: Xenwatch Multithreading - Dongli Zhang, Oracle
 
Enabling POWER 8 advanced features on Linux
Enabling POWER 8 advanced features on LinuxEnabling POWER 8 advanced features on Linux
Enabling POWER 8 advanced features on Linux
 
XPDDS18: The Art of Virtualizing Cache Maintenance - Julien Grall, Arm
XPDDS18: The Art of Virtualizing Cache Maintenance - Julien Grall, ArmXPDDS18: The Art of Virtualizing Cache Maintenance - Julien Grall, Arm
XPDDS18: The Art of Virtualizing Cache Maintenance - Julien Grall, Arm
 
XPDDS18: Memory Overcommitment in XEN - Huang Zhichao, Huawei
XPDDS18: Memory Overcommitment in XEN - Huang Zhichao, HuaweiXPDDS18: Memory Overcommitment in XEN - Huang Zhichao, Huawei
XPDDS18: Memory Overcommitment in XEN - Huang Zhichao, Huawei
 
Performance: Observe and Tune
Performance: Observe and TunePerformance: Observe and Tune
Performance: Observe and Tune
 
TDS-16489U-R2 0215 EN
TDS-16489U-R2 0215 ENTDS-16489U-R2 0215 EN
TDS-16489U-R2 0215 EN
 
Keith Paskett - Postgres on ZFS @ Postgres Open
Keith Paskett - Postgres on ZFS @ Postgres OpenKeith Paskett - Postgres on ZFS @ Postgres Open
Keith Paskett - Postgres on ZFS @ Postgres Open
 
Ibm aix technical deep dive workshop advanced administration and problem dete...
Ibm aix technical deep dive workshop advanced administration and problem dete...Ibm aix technical deep dive workshop advanced administration and problem dete...
Ibm aix technical deep dive workshop advanced administration and problem dete...
 
XPDS16: Xen Live Patching - Updating Xen Without Rebooting - Konrad Wilk, Ora...
XPDS16: Xen Live Patching - Updating Xen Without Rebooting - Konrad Wilk, Ora...XPDS16: Xen Live Patching - Updating Xen Without Rebooting - Konrad Wilk, Ora...
XPDS16: Xen Live Patching - Updating Xen Without Rebooting - Konrad Wilk, Ora...
 
XPDDS17: Xen-lite for ARM: Adapting Xen for a Samsung Exynos MicroServer with...
XPDDS17: Xen-lite for ARM: Adapting Xen for a Samsung Exynos MicroServer with...XPDDS17: Xen-lite for ARM: Adapting Xen for a Samsung Exynos MicroServer with...
XPDDS17: Xen-lite for ARM: Adapting Xen for a Samsung Exynos MicroServer with...
 
제3회난공불락 오픈소스 인프라세미나 - lustre
제3회난공불락 오픈소스 인프라세미나 - lustre제3회난공불락 오픈소스 인프라세미나 - lustre
제3회난공불락 오픈소스 인프라세미나 - lustre
 
XPDDS17: How to Abstract Hardware Acceleration Device in Cloud Environment - ...
XPDDS17: How to Abstract Hardware Acceleration Device in Cloud Environment - ...XPDDS17: How to Abstract Hardware Acceleration Device in Cloud Environment - ...
XPDDS17: How to Abstract Hardware Acceleration Device in Cloud Environment - ...
 
Design decision nfs-versus_fc_storage v_0.3
Design decision nfs-versus_fc_storage v_0.3Design decision nfs-versus_fc_storage v_0.3
Design decision nfs-versus_fc_storage v_0.3
 
Relax and Recover on POWER (Updated 05-2017)
Relax and Recover on POWER (Updated 05-2017)Relax and Recover on POWER (Updated 05-2017)
Relax and Recover on POWER (Updated 05-2017)
 
KVM Tuning @ eBay
KVM Tuning @ eBayKVM Tuning @ eBay
KVM Tuning @ eBay
 
IMCSummit 2015 - Day 2 Developer Track - The NVM Revolution
IMCSummit 2015 - Day 2 Developer Track - The NVM RevolutionIMCSummit 2015 - Day 2 Developer Track - The NVM Revolution
IMCSummit 2015 - Day 2 Developer Track - The NVM Revolution
 
XPDDS18: NVDIMM Overview - George Dunlap, Citrix
XPDDS18: NVDIMM Overview - George Dunlap, Citrix XPDDS18: NVDIMM Overview - George Dunlap, Citrix
XPDDS18: NVDIMM Overview - George Dunlap, Citrix
 
XPDS13: Perf Support in Xen - Boris Ostrovsky, Oracle
XPDS13: Perf Support in Xen - Boris Ostrovsky, OracleXPDS13: Perf Support in Xen - Boris Ostrovsky, Oracle
XPDS13: Perf Support in Xen - Boris Ostrovsky, Oracle
 

Similar to Virtual SQL Servers. Actual Performance.

Re-Think Storage – PernixData. Meet & greet with Frank Denneman
Re-Think Storage – PernixData. Meet & greet with Frank DennemanRe-Think Storage – PernixData. Meet & greet with Frank Denneman
Re-Think Storage – PernixData. Meet & greet with Frank Denneman
Digicomp Academy AG
 
Objective SQL Server Performance
Objective SQL Server PerformanceObjective SQL Server Performance
Objective SQL Server Performance
David Klee
 
Sharing High-Performance Interconnects Across Multiple Virtual Machines
Sharing High-Performance Interconnects Across Multiple Virtual MachinesSharing High-Performance Interconnects Across Multiple Virtual Machines
Sharing High-Performance Interconnects Across Multiple Virtual Machines
inside-BigData.com
 
Achieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVMAchieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVM
DevOps.com
 
Achieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVMAchieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVM
data://disrupted®
 
VMworld 2013: How SRP Delivers More Than Power to Their Customers
VMworld 2013: How SRP Delivers More Than Power to Their Customers VMworld 2013: How SRP Delivers More Than Power to Their Customers
VMworld 2013: How SRP Delivers More Than Power to Their Customers
VMworld
 
Accelerating Hyper-Converged Enterprise Virtualization using Proxmox and Ceph
Accelerating Hyper-Converged Enterprise Virtualization using Proxmox and CephAccelerating Hyper-Converged Enterprise Virtualization using Proxmox and Ceph
Accelerating Hyper-Converged Enterprise Virtualization using Proxmox and Ceph
Bangladesh Network Operators Group
 
EMC Vmax3 tech-deck deep dive
EMC Vmax3 tech-deck deep diveEMC Vmax3 tech-deck deep dive
EMC Vmax3 tech-deck deep dive
solarisyougood
 
Platform Security Summit 18: Xen Security Weather Report 2018
Platform Security Summit 18: Xen Security Weather Report 2018Platform Security Summit 18: Xen Security Weather Report 2018
Platform Security Summit 18: Xen Security Weather Report 2018
The Linux Foundation
 
Best Practices for Virtualizing Apache Hadoop
Best Practices for Virtualizing Apache HadoopBest Practices for Virtualizing Apache Hadoop
Best Practices for Virtualizing Apache Hadoop
Hortonworks
 
Ceph Day Beijing - Ceph all-flash array design based on NUMA architecture
Ceph Day Beijing - Ceph all-flash array design based on NUMA architectureCeph Day Beijing - Ceph all-flash array design based on NUMA architecture
Ceph Day Beijing - Ceph all-flash array design based on NUMA architecture
Ceph Community
 
Ceph Day Beijing - Ceph All-Flash Array Design Based on NUMA Architecture
Ceph Day Beijing - Ceph All-Flash Array Design Based on NUMA ArchitectureCeph Day Beijing - Ceph All-Flash Array Design Based on NUMA Architecture
Ceph Day Beijing - Ceph All-Flash Array Design Based on NUMA Architecture
Danielle Womboldt
 
Benchmark emc vnx7500, emc fast suite, emc snap sure and oracle rac on v-mware
Benchmark   emc vnx7500, emc fast suite, emc snap sure and oracle rac on v-mwareBenchmark   emc vnx7500, emc fast suite, emc snap sure and oracle rac on v-mware
Benchmark emc vnx7500, emc fast suite, emc snap sure and oracle rac on v-mware
solarisyougood
 
Advanced performance troubleshooting using esxtop
Advanced performance troubleshooting using esxtopAdvanced performance troubleshooting using esxtop
Advanced performance troubleshooting using esxtop
Alan Renouf
 
Accelerating Cassandra Workloads on Ceph with All-Flash PCIE SSDS
Accelerating Cassandra Workloads on Ceph with All-Flash PCIE SSDSAccelerating Cassandra Workloads on Ceph with All-Flash PCIE SSDS
Accelerating Cassandra Workloads on Ceph with All-Flash PCIE SSDS
Ceph Community
 
Vx Rack : L'hyperconvergence avec l'experience VCE
Vx Rack : L'hyperconvergence avec l'experience VCEVx Rack : L'hyperconvergence avec l'experience VCE
Vx Rack : L'hyperconvergence avec l'experience VCE
RSD
 
Polyteda Power DRC/LVS July 2016
Polyteda Power DRC/LVS July 2016Polyteda Power DRC/LVS July 2016
Polyteda Power DRC/LVS July 2016
Oleksandra Nazola
 
Emc vspex customer_presentation_private_cloud_virtualized_share_point
Emc vspex customer_presentation_private_cloud_virtualized_share_pointEmc vspex customer_presentation_private_cloud_virtualized_share_point
Emc vspex customer_presentation_private_cloud_virtualized_share_point
xKinAnx
 
LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...
LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...
LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...
The Linux Foundation
 
Install FD.IO VPP On Intel(r) Architecture & Test with Trex*
Install FD.IO VPP On Intel(r) Architecture & Test with Trex*Install FD.IO VPP On Intel(r) Architecture & Test with Trex*
Install FD.IO VPP On Intel(r) Architecture & Test with Trex*
Michelle Holley
 

Similar to Virtual SQL Servers. Actual Performance. (20)

Re-Think Storage – PernixData. Meet & greet with Frank Denneman
Re-Think Storage – PernixData. Meet & greet with Frank DennemanRe-Think Storage – PernixData. Meet & greet with Frank Denneman
Re-Think Storage – PernixData. Meet & greet with Frank Denneman
 
Objective SQL Server Performance
Objective SQL Server PerformanceObjective SQL Server Performance
Objective SQL Server Performance
 
Sharing High-Performance Interconnects Across Multiple Virtual Machines
Sharing High-Performance Interconnects Across Multiple Virtual MachinesSharing High-Performance Interconnects Across Multiple Virtual Machines
Sharing High-Performance Interconnects Across Multiple Virtual Machines
 
Achieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVMAchieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVM
 
Achieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVMAchieving the Ultimate Performance with KVM
Achieving the Ultimate Performance with KVM
 
VMworld 2013: How SRP Delivers More Than Power to Their Customers
VMworld 2013: How SRP Delivers More Than Power to Their Customers VMworld 2013: How SRP Delivers More Than Power to Their Customers
VMworld 2013: How SRP Delivers More Than Power to Their Customers
 
Accelerating Hyper-Converged Enterprise Virtualization using Proxmox and Ceph
Accelerating Hyper-Converged Enterprise Virtualization using Proxmox and CephAccelerating Hyper-Converged Enterprise Virtualization using Proxmox and Ceph
Accelerating Hyper-Converged Enterprise Virtualization using Proxmox and Ceph
 
EMC Vmax3 tech-deck deep dive
EMC Vmax3 tech-deck deep diveEMC Vmax3 tech-deck deep dive
EMC Vmax3 tech-deck deep dive
 
Platform Security Summit 18: Xen Security Weather Report 2018
Platform Security Summit 18: Xen Security Weather Report 2018Platform Security Summit 18: Xen Security Weather Report 2018
Platform Security Summit 18: Xen Security Weather Report 2018
 
Best Practices for Virtualizing Apache Hadoop
Best Practices for Virtualizing Apache HadoopBest Practices for Virtualizing Apache Hadoop
Best Practices for Virtualizing Apache Hadoop
 
Ceph Day Beijing - Ceph all-flash array design based on NUMA architecture
Ceph Day Beijing - Ceph all-flash array design based on NUMA architectureCeph Day Beijing - Ceph all-flash array design based on NUMA architecture
Ceph Day Beijing - Ceph all-flash array design based on NUMA architecture
 
Ceph Day Beijing - Ceph All-Flash Array Design Based on NUMA Architecture
Ceph Day Beijing - Ceph All-Flash Array Design Based on NUMA ArchitectureCeph Day Beijing - Ceph All-Flash Array Design Based on NUMA Architecture
Ceph Day Beijing - Ceph All-Flash Array Design Based on NUMA Architecture
 
Benchmark emc vnx7500, emc fast suite, emc snap sure and oracle rac on v-mware
Benchmark   emc vnx7500, emc fast suite, emc snap sure and oracle rac on v-mwareBenchmark   emc vnx7500, emc fast suite, emc snap sure and oracle rac on v-mware
Benchmark emc vnx7500, emc fast suite, emc snap sure and oracle rac on v-mware
 
Advanced performance troubleshooting using esxtop
Advanced performance troubleshooting using esxtopAdvanced performance troubleshooting using esxtop
Advanced performance troubleshooting using esxtop
 
Accelerating Cassandra Workloads on Ceph with All-Flash PCIE SSDS
Accelerating Cassandra Workloads on Ceph with All-Flash PCIE SSDSAccelerating Cassandra Workloads on Ceph with All-Flash PCIE SSDS
Accelerating Cassandra Workloads on Ceph with All-Flash PCIE SSDS
 
Vx Rack : L'hyperconvergence avec l'experience VCE
Vx Rack : L'hyperconvergence avec l'experience VCEVx Rack : L'hyperconvergence avec l'experience VCE
Vx Rack : L'hyperconvergence avec l'experience VCE
 
Polyteda Power DRC/LVS July 2016
Polyteda Power DRC/LVS July 2016Polyteda Power DRC/LVS July 2016
Polyteda Power DRC/LVS July 2016
 
Emc vspex customer_presentation_private_cloud_virtualized_share_point
Emc vspex customer_presentation_private_cloud_virtualized_share_pointEmc vspex customer_presentation_private_cloud_virtualized_share_point
Emc vspex customer_presentation_private_cloud_virtualized_share_point
 
LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...
LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...
LCNA14: Why Use Xen for Large Scale Enterprise Deployments? - Konrad Rzeszute...
 
Install FD.IO VPP On Intel(r) Architecture & Test with Trex*
Install FD.IO VPP On Intel(r) Architecture & Test with Trex*Install FD.IO VPP On Intel(r) Architecture & Test with Trex*
Install FD.IO VPP On Intel(r) Architecture & Test with Trex*
 

Recently uploaded

Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...
Product School
 
Elevating Tactical DDD Patterns Through Object Calisthenics
Elevating Tactical DDD Patterns Through Object CalisthenicsElevating Tactical DDD Patterns Through Object Calisthenics
Elevating Tactical DDD Patterns Through Object Calisthenics
Dorra BARTAGUIZ
 
Connector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a buttonConnector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a button
DianaGray10
 
Accelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish CachingAccelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish Caching
Thijs Feryn
 
Epistemic Interaction - tuning interfaces to provide information for AI support
Epistemic Interaction - tuning interfaces to provide information for AI supportEpistemic Interaction - tuning interfaces to provide information for AI support
Epistemic Interaction - tuning interfaces to provide information for AI support
Alan Dix
 
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
UiPathCommunity
 
Bits & Pixels using AI for Good.........
Bits & Pixels using AI for Good.........Bits & Pixels using AI for Good.........
Bits & Pixels using AI for Good.........
Alison B. Lowndes
 
Essentials of Automations: Optimizing FME Workflows with Parameters
Essentials of Automations: Optimizing FME Workflows with ParametersEssentials of Automations: Optimizing FME Workflows with Parameters
Essentials of Automations: Optimizing FME Workflows with Parameters
Safe Software
 
PCI PIN Basics Webinar from the Controlcase Team
PCI PIN Basics Webinar from the Controlcase TeamPCI PIN Basics Webinar from the Controlcase Team
PCI PIN Basics Webinar from the Controlcase Team
ControlCase
 
JMeter webinar - integration with InfluxDB and Grafana
JMeter webinar - integration with InfluxDB and GrafanaJMeter webinar - integration with InfluxDB and Grafana
JMeter webinar - integration with InfluxDB and Grafana
RTTS
 
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Jeffrey Haguewood
 
GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
Sri Ambati
 
GraphRAG is All You need? LLM & Knowledge Graph
GraphRAG is All You need? LLM & Knowledge GraphGraphRAG is All You need? LLM & Knowledge Graph
GraphRAG is All You need? LLM & Knowledge Graph
Guy Korland
 
Empowering NextGen Mobility via Large Action Model Infrastructure (LAMI): pav...
Empowering NextGen Mobility via Large Action Model Infrastructure (LAMI): pav...Empowering NextGen Mobility via Large Action Model Infrastructure (LAMI): pav...
Empowering NextGen Mobility via Large Action Model Infrastructure (LAMI): pav...
Thierry Lestable
 
FIDO Alliance Osaka Seminar: FIDO Security Aspects.pdf
FIDO Alliance Osaka Seminar: FIDO Security Aspects.pdfFIDO Alliance Osaka Seminar: FIDO Security Aspects.pdf
FIDO Alliance Osaka Seminar: FIDO Security Aspects.pdf
FIDO Alliance
 
Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*
Frank van Harmelen
 
Key Trends Shaping the Future of Infrastructure.pdf
Key Trends Shaping the Future of Infrastructure.pdfKey Trends Shaping the Future of Infrastructure.pdf
Key Trends Shaping the Future of Infrastructure.pdf
Cheryl Hung
 
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
Product School
 
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Product School
 
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
Tobias Schneck
 

Recently uploaded (20)

Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...
 
Elevating Tactical DDD Patterns Through Object Calisthenics
Elevating Tactical DDD Patterns Through Object CalisthenicsElevating Tactical DDD Patterns Through Object Calisthenics
Elevating Tactical DDD Patterns Through Object Calisthenics
 
Connector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a buttonConnector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a button
 
Accelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish CachingAccelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish Caching
 
Epistemic Interaction - tuning interfaces to provide information for AI support
Epistemic Interaction - tuning interfaces to provide information for AI supportEpistemic Interaction - tuning interfaces to provide information for AI support
Epistemic Interaction - tuning interfaces to provide information for AI support
 
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
 
Bits & Pixels using AI for Good.........
Bits & Pixels using AI for Good.........Bits & Pixels using AI for Good.........
Bits & Pixels using AI for Good.........
 
Essentials of Automations: Optimizing FME Workflows with Parameters
Essentials of Automations: Optimizing FME Workflows with ParametersEssentials of Automations: Optimizing FME Workflows with Parameters
Essentials of Automations: Optimizing FME Workflows with Parameters
 
PCI PIN Basics Webinar from the Controlcase Team
PCI PIN Basics Webinar from the Controlcase TeamPCI PIN Basics Webinar from the Controlcase Team
PCI PIN Basics Webinar from the Controlcase Team
 
JMeter webinar - integration with InfluxDB and Grafana
JMeter webinar - integration with InfluxDB and GrafanaJMeter webinar - integration with InfluxDB and Grafana
JMeter webinar - integration with InfluxDB and Grafana
 
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
 
GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
 
GraphRAG is All You need? LLM & Knowledge Graph
GraphRAG is All You need? LLM & Knowledge GraphGraphRAG is All You need? LLM & Knowledge Graph
GraphRAG is All You need? LLM & Knowledge Graph
 
Empowering NextGen Mobility via Large Action Model Infrastructure (LAMI): pav...
Empowering NextGen Mobility via Large Action Model Infrastructure (LAMI): pav...Empowering NextGen Mobility via Large Action Model Infrastructure (LAMI): pav...
Empowering NextGen Mobility via Large Action Model Infrastructure (LAMI): pav...
 
FIDO Alliance Osaka Seminar: FIDO Security Aspects.pdf
FIDO Alliance Osaka Seminar: FIDO Security Aspects.pdfFIDO Alliance Osaka Seminar: FIDO Security Aspects.pdf
FIDO Alliance Osaka Seminar: FIDO Security Aspects.pdf
 
Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*
 
Key Trends Shaping the Future of Infrastructure.pdf
Key Trends Shaping the Future of Infrastructure.pdfKey Trends Shaping the Future of Infrastructure.pdf
Key Trends Shaping the Future of Infrastructure.pdf
 
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
 
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...
 
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
 

Virtual SQL Servers. Actual Performance.

  • 1.
  • 2. © Heraflux Technologies® 2 @kleegeek davidklee.net heraflux.com linkedin.com/in/davidaklee Specialties / Focus Areas / Passions: • Performance Tuning & Troubleshooting • Virtualization • Cloud Enablement • Infrastructure Architecture • Health & Efficiency • Capacity Management Founder & Chief Architect
  • 3.  What is Virtualization?  Resources and queues  Physical servers  Storage and interconnects  VM architecture  Inside the VM © Heraflux Technologies® 3
  • 4.  Added layer  Physical resources  Resource queues  Very small queue delays  When done right…  Shared everything architecture © Heraflux Technologies® 4
  • 5. 150 GHz CPU 4 TB Memory 4x10GbE Network 20 TB Tier 1 Storage 40 TB Tier 2 Storage VM 16 vCPU 128 GB vRAM VM 8 vCPU 64 GB vRAM VM 2 vCPU 16 GB vRAM VM 2 vCPU 16 GB vRAM VM 2 vCPU 16 GB vRAM VM 2 vCPU 16 GB vRAM VM 2 vCPU 16 GB vRAM VM 2 vCPU 16 GB vRAM V I R T U A L I Z A T I O N © Heraflux Technologies® 5
  • 6. Hypervisor CPU Scheduler CPU Execution CPU Scheduling Queue Memory Allocator Mem R / W Mem Allocation Queue Disk Scheduler Disk R / W Disk Scheduling Queue Network Scheduler Network Tran / Rec Network Scheduling Queue VM TASK VM TASK VM TASK VM TASK VM TASK © Heraflux Technologies® 6
  • 7. • Finite! • Single compute node hardware • Total cluster compute capacity • Storage speed (IOPs, throughput) • VM maximums • Interconnect path speed Hard Limits (Resources) Soft Limits (Queues) • Measurable but not finite • Memory oversubscription • CPU scheduler contention • Shared resource utilization • “Noisy Neighbors” © Heraflux Technologies® 7• 7 • 7
  • 8. © Heraflux Technologies® 8 Networking Interconnects Storage Physical Server Virtualization Operating System SQL Server Instance SQL Server DB Application
  • 10.  Most shared  Most critical  Most complex  Most problematic  Slowest piece of the stack  Many individual points of contention © Heraflux Technologies® 10
  • 11. © Heraflux Technologies® 11 Controller Controller LUN LUN LUN LUN Disk Pool T1 T2 T3
  • 12.  Test raw performance  SQLIO Deprecated!  DiskSpd Batch heraflux.com/go/diskspd  Collect metrics:  I/Os per second (IOPs)  Latency (ms)  Throughput (MB/s) 0.00 10000.00 20000.00 30000.00 40000.00 50000.00 60000.00 70000.00 1 2 4 8 16 32 64 128 IOps Thread Intensity IOps Per Operations per Thread Sequential Read Random Read Sequential Write Random Write © Heraflux Technologies® 12
  • 15.  Manufacturer  Intel  AMD  Sockets  Cores  Speed vs. GHz  Logical vs. Physical  Hyperthreading © Heraflux Technologies® 15
  • 16. © Heraflux Technologies® 16 (Img src: http://frankdenneman.nl/2011/01/05/amd-magny-cours-and-esx/) CPU CPU CPU CPU Memory
  • 17. © Heraflux Technologies® 17 (Img src: http://www.learnyourtech.com/hardware/)
  • 18.  Sample Cisco UCS memory config  Slot & chip placement © Heraflux Technologies® 18 (Src: http://www.cisco.com/c/dam/en/us/products/collateral/servers-unified-computing/ucs-b-series-blade-servers/b200m4-specsheet.pdf page 44)
  • 20.  Resource limits are easy to detect / work around  Queue contention much harder  Time in queue = time lost from VM  Silent performance killer  Everything in a VM must be scheduled  … including idle resources  Queue processing is not always FIFO © Heraflux Technologies® 20
  • 21.  Get physical machine configuration  Try to fit VM inside one NUMA node  Otherwise, balance across number of NUMA nodes  Test configurations for best results © Heraflux Technologies® 21
  • 22.  Example: 16 vCPU VM  What’s better?  2 vSocket x 8 vCore?  4 vSocket x 4 vCore?  8 vSocket x 2 vCore?  Varies by workload, hardware  Test it for yourself! © Heraflux Technologies® 22 0 100000 200000 300000 400000 500000 600000 700000 800000 900000 8 16 64 256 Transactions/min Concurrent HammerDB Users vNUMA SQL Server Scalability - 16 vCPUs - HammerDB 4socket x 4CPU 8socket x 2CPU 2socket x 8CPU
  • 23.  1 vCPU to 1 pCPU is a poor recommendation  Queues  Ready Time  Co-Stop  Ex: Client environment  74 running VMs  VMs @ 2-8 vCPU  72 total pCores  Fantastic performance © Heraflux Technologies® 23
  • 27.  “Right” amount of vCPU and vRAM resources  Physical world = Size for end of life  Virtual world = Size for right now  Idle vCPUs will slow application VMs performance  Repeat “right-sizing” analysis routinely © Heraflux Technologies® 27
  • 28.  Virtual disks whenever possible  Multiple SCSI controllers  VMware – PVSCSI  Spread out the workload © Heraflux Technologies® 28
  • 29.  C: - Operating System  D: - SQL Server Instance Home  E: - System Databases (master, model, msdb) *  F: - User Database Data (1 of X)…  G: - User Database Log (1 of Y)…  H: - TempDB…  Y: - Windows Page file **  Z: - Backups Adjust as necessary (but stay standardized) © Heraflux Technologies® 29
  • 30.  64KB NTFS block sizes  Set power settings to “High Performance” (CPU-Z)  Set antivirus exclusions for SQL Server (tinyurl.com/sqlav)  Ongoing OS-level performance metric collection  No greater than five minute interval  Windows Perfmon, Microsoft SCOM, or other third-party utility  heraflux.com/go/perfmon © Heraflux Technologies® 30
  • 31.  Goal: Maximize performance while reducing resource scheduling  Parallelizable workloads  Determine how parallel the workload is  Set MaxDOP = vNUMA node core count (?)  Cost threshold for parallelism = Not default  Jonathan Kehayias – Tuning CTOP from Plan Cache  bit.ly/1rTs9UX © Heraflux Technologies® 31
  • 32.  Spread out the I/O  File groups, data files, partitions  Parallelism with multiple active storage paths  Reduce I/O  Table / index compression vs. SAN compression  In-memory constructs  More RAM  SSD read / write caching  Faster I/O – All flash SAN  And then clean up bad schemas & queries! © Heraflux Technologies® 32
  • 33.  Virtualization works.  Equivalent performance if done right  Efficiency in data handling © Heraflux Technologies® 33
  • 34. © Heraflux Technologies® 34 @kleegeek davidklee.net heraflux.com linkedin.com/in/davidaklee

Editor's Notes

  1. Start with the coffee cup / vnx fiber issue story
  2. SQLIO Batch