SPECstorage(TM) Solution 2020_eda_blended Result Microsoft and NetApp : Azure NetApp Files large volume breakthrough mode (Aug Inc. 2026) SPECstorage Solution = 5340 Job_Sets (Overall Response Time = 0.54 msec) 2020_eda_blended =============================================================================== Performance =========== Business Average Metric Latency Job_Sets Job_Sets (Job_Sets) (msec) Ops/Sec MB/Sec ------------ ------------ ------------ ------------ 356 0.3 160208 2584 712 0.3 320416 5168 1068 0.3 480626 7754 1424 0.3 640836 10340 1780 0.3 801044 12926 2136 0.3 961253 15508 2492 0.3 1121460 18093 2848 0.4 1281671 20678 3204 0.4 1441877 23265 3560 0.5 1602089 25850 3916 0.6 1762298 28438 4272 0.7 1922506 31024 4628 0.9 2082716 33605 4984 1.2 2242898 36189 5340 2.0 2403131 38776 =============================================================================== Product and Test Information ============================ +---------------------------------------------------------------+ | Azure NetApp Files large volume breakthrough mode (Aug 2026) | +---------------------------------------------------------------+ Tested by Microsoft and NetApp Inc. Hardware Available August 2026 Software Available May 2026 Date Tested August 2026 License Number 33 Licensee Locations San Jose, CA USA Azure NetApp Files is an Azure native, first-party, enterprise-class, high-performance file storage service. It provides Volumes as a service, which you can create within a NetApp account and a capacity pool, and share to clients using SMB and NFS. You can also select service and performance levels and manage data protection. You can create and manage high-performance, highly available, and scalable file shares by using the same protocols and tools that you're familiar with and rely on on-premises. Solution Under Test Bill of Materials ===================================== Item No Qty Type Vendor Model/Name Description ---- ---- ---------- ---------- ---------- ----------------------------------- 1 1 Storage Microsoft Azure Azure NetApp Files large volumes can volume bre support from 50 TiB to 2 PiB in akthrough size. Breakthrough mode enables mode individual volumes to scale beyond 12.5 GiB/s. Volumes can be resized up or down on demand, and throughput can be adjusted automatically (based on volume size) or manually depending on the capacity pool QoS type. 2 48 Azure Microsoft Standard_E Red Hat Enterprise Linux running on Virtual 8_v5 48 Azure E8_v5 Virtual Machines (8 Machine vCPU, 64 GiB Memory, 1x Virtual NIC supporting up to 12.5 Gbps egress network bandwidth). Each workload client mounted the Azure NetApp Files large volume once over NFSv3. 3 1 Azure Microsoft Standard_D (Prime Client) Red Hat Enterprise Virtual 2s_v5 Linux running on an Azure D2s_v5 Machine Virtual Machine (2 vCPU, 8 GiB Memory, 1x Virtual NIC supporting up to 12.5 Gbps egress network bandwidth). Configuration Diagrams ====================== 1) storage2020-20260901-00150.config1.png (see SPECstorage Solution 2020 results webpage) Component Software ================== Item Name and No Component Type Version Description ---- ------------ ------------ ------------ ----------------------------------- 1 RHEL 9.8 Operating RHEL 9.8, Operating System (OS) for the System kernel 5.14. workload clients 0-687.41.1.e l9_8.x86_64 Hardware Configuration and Tuning - Virtual =========================================== +----------------------------------------------------------------------+ | Client Network Settings | +----------------------------------------------------------------------+ Parameter Name Value Description --------------- --------------- ---------------------------------------- Accelerated Enabled Accelerated Networking enables single Networking root I/O virtualization (SR-IOV) on supported virtual machine (VM) types +----------------------------------------------------------------------+ | Storage Network Settings | +----------------------------------------------------------------------+ Parameter Name Value Description --------------- --------------- ---------------------------------------- Network Standard Standard Network Features allows Azure features VNet features such as network security groups, user-defined routes and others. Hardware Configuration and Tuning Notes --------------------------------------- None Software Configuration and Tuning - Virtual =========================================== +----------------------------------------------------------------------+ | Clients | +----------------------------------------------------------------------+ Parameter Name Value Description --------------- --------------- ---------------------------------------- rsize,wsize 262144 NFS mount options for data block size protocol tcp NFS mount options for protocol vers 3 NFS mount options for NFS version nconnect 8 NFS mount options for multiple TCP connections actimeo 600 NFS mount option to modify the timeouts for attribute caching nocto present NFS mount option to turn off close-to- (boolean) open consistency noatime present NFS mount option to turn off access time (boolean) updates nofile 102400 Maximum number of open files per user nproc 10240 Maximum number of processes per user sunrpc.tcp_slot 128 Sets the number of (TCP) RPC entries to _table_entries pre-allocate for in-flight RPC requests net.core.wmem_m 16777216 Maximum size of the socket send buffer ax net.core.rmem_m 16777216 Maximum size of the socket receive ax buffer net.core.wmem_d 1048576 Default setting in bytes of the socket efault send buffer net.core.rmem_d 1048576 Default setting in bytes of the socket efault receive buffer net.ipv4.tcp_rm 1048576 8388608 Minimum, default and maximum size of the em 33554432 TCP receive buffer net.ipv4.tcp_wm 1048576 8388608 Minimum, default and maximum size of the em 33554432 TCP send buffer net.core.optmem 4194304 Maximum ancillary buffer size allowed _max per socket net.core.somaxc 65535 Maximum tcp backlog an application can onn request net.ipv4.tcp_me 4096 89600 Maximum memory in 4096-byte pages across m 8388608 all TCP applications. Contains minimum, pressure and maximum. net.ipv4.tcp_wi 1 Enable TCP window scaling ndow_scaling net.ipv4.tcp_ti 0 Turn off timestamps to reduce mestamps performance spikes related to timestamp generation net.ipv4.tcp_no 1 Prevent TCP from caching connection _metrics_save metrics on closing connections net.ipv4.route. 1 Flush the routing cache flush net.ipv4.tcp_lo 1 Allows TCP to make decisions to prefer w_latency lower latency instead of maximizing network throughput. * See Note below net.ipv4.ip_loc 1024 65000 Defines the local port range that is al_port_range used by TCP and UDP traffic to choose the local port. net.ipv4.tcp_sl 0 Congestion window will not be timed out ow_start_after_ after an idle period idle net.core.netdev 300000 Sets maximum number of packets, queued _max_backlog on the input side, when the interface receives packets faster than kernel can process net.ipv4.tcp_sa 0 Disable TCP selective acknowledgements ck net.ipv4.tcp_ds 0 Disable duplicate SACKs ack net.ipv4.tcp_fa 0 Disable forward acknowledgement. * See ck Note below vm.dirty_expire 30000 Defines when dirty data is old enough to _centisecs be eligible for writeout by the kernel flusher threads. Unit is 100ths of a second. vm.dirty_writeb 30000 Defines a time interval between periodic ack_centisecs wake-ups of the kernel threads responsible for writing dirty data to hard-disk. Unit is 100ths of a second. Software Configuration and Tuning Notes --------------------------------------- Tuned the necessary client parameters as shown above, for communication between clients and storage over Azure Virtual Networking, to optimize data transfer and minimize overhead. < br /> Note - net.ipv4.tcp_low_latency and net.ipv4.tcp_fack are obsolete in Linux 4.14 and later but were configured during the test. Service SLA Notes ----------------- Service Level Agreement (SLA) for Azure NetApp Files Storage and Filesystems ======================= Item Stable No Description Data Protection Storage Qty ---- ------------------------------------- ------------------ -------- ----- 1 Azure NetApp Files large volume, Azure NetApp Files Stable 1 Flexible Service Level, 150 TiB, Flexible, Storage 57,600 provisioned MiB/s Standard, Premium and Ultra service levels are built on a fault- tolerant bare- metal fleet powered by ONTAP, delivering enterprise-grade resilience, and uses RAID-DP (Double Parity RAID) to safeguard data against disk failures. This mechanism distributes parity across multiple disks, enabling seamless data recovery even if two disks fail simultaneously. RAID-DP has a long-standing presence in the enterprise storage industry and is recognized for its proven reliability and fault tolerance. Number of Filesystems 1 Total Capacity 150 TiB Filesystem Type Azure NetApp Files large volume Filesystem Creation Notes ------------------------- The large volume was created via the public Azure API using the Azure CLI. Creation commands are available here: https://learn.microsoft.com/en-us/cli/azure/netappfiles/volume?view=azure-cli-latest#az-netappfiles-volume-create
Creating the Azure NetApp Files Account: az netappfiles account create --account-name [account-name] --resource-group [resource-group] --location [location]
Creating the Azure NetApp Files Capacity Pool: az netappfiles pool create --account-name [account-name] --resource-group [resource-group] --location [location] --pool-name [pool-name] --service-level Flexible --size 164926744166400 --CustomThroughputMibps 57600
Creating the Azure NetApp Files Volume: az netappfiles volume create --account-name [account-name] --file-path [mount-point] --name [volume-name] --pool-name [pool-name] --resource-group [resource-group] --location [location] --protocol-types NFSv3 --vnet [vnet-id] --subnet [subnet-id] --zones 1 --service-level Flexible --network-features Standard --is-large-volume true --usage-threshold 153600 --throughput-mibps 57600 --breakthrough-mode Enabled
Storage and Filesystem Notes ---------------------------- The provisioned throughput (57,600 MiB/s) is the volume's configured maximum; actual maximum achievable throughput depends on the workload. Transport Configuration - Virtual ================================= Item Number of No Transport Type Ports Used Notes ---- --------------- ---------- ----------------------------------------------- 1 Virtual NIC 48 Each of the 48 workload Linux Virtual Machines has a single Network Adapter supporting up to 12.5 Gbps egress network bandwidth with Accelerated Networking enabled. 2 Virtual NIC 1 The Prime client has a single Network Adapter supporting up to 12.5 Gbps egress network bandwidth with Accelerated Networking enabled. Transport Configuration Notes ----------------------------- Azure Virtual machines bandwidth limits egress (outbound) traffic from each virtual machine up to 12.5 Gbps. The virtual machines ingress bandwidth rates may exceed the documented limit depending on other resources available to the virtual machine (https://learn.microsoft.com/en-us/azure/virtual-network/virtual-machine-network-throughput) Switches - Virtual ================== Total Used Item Port Port No Switch Name Switch Type Count Count Notes ---- -------------------- --------------- ------ ----- ------------------------ 1 Azure Virtual Virtual Network 55 55 The Azure virtual Network - Sweden network had 6 Central region - AZ1 connections for the Azure NetApp Files storage endpoints, 1 connection for the Prime client, and 48 (1 per) RHEL client VMs. Azure virtual networks allow up to 65,536 Network interface cards and Private IP addresses per virtual network. Processing Elements - Virtual ============================= Item No Qty Type Location Description Processing Function ---- ---- -------- -------------- ------------------------- ------------------- 1 384 vCPU Azure Cloud Intel(R) Xeon(R) Platinum Client Workload 8573C CPU @ 3.50GHz (8 Generator cores allocated to each of the 48 Workload VMs) 2 2 vCPU Azure Cloud Intel(R) Xeon(R) Platinum Prime Client 8370C CPU @ 2.80GHz (2 cores allocated to Prime VM) Processing Element Notes ------------------------ n/a Memory - Virtual ================ Size in Number of Description GiB Instances Nonvolatile Total GiB ------------------------- ---------- ---------- ------------ ------------ Client Workload Generator 64 48 V 3072 Prime Client 8 1 V 8 Grand Total Memory Gibibytes 3080 Memory Notes ------------ n/a Stable Storage ============== Azure NetApp Files utilizes non-volatile battery-backed memory of two independent nodes as write caching prior to write acknowledgement. This protects the filesystem from any single-point-of-failure until the data is de-staged to disks. In the event of an abrupt failure, pending data in the non-volatile battery-backed memory is replayed to disk upon restoration. Solution Under Test Configuration Notes ======================================= Each client was round-robin assigned and mounted NFSv3 to one of six storage endpoints for the Azure NetApp Files large volume. Each storage endpoint has 8 workload VMs connected.
Unlike a general-purpose operating system, Azure NetApp Files does not provide mechanisms for customers to run third-party code (https://learn.microsoft.com/en-us/security/benchmark/azure/baselines/azure-netapp-files-security-baseline?toc=/azure/azure-netapp-files/TOC.json#security-profile). Azure Resource Manager allows only an allow-listed set of operations to be executed via the Azure APIs (https://learn.microsoft.com/en-us/azure/azure-netapp-files/control-plane-security).
Underlying Azure infrastructure was patched for Spectre/Meltdown on or prior to January 2018. (https://azure.microsoft.com/en-us/blog/securing-azure-customers-from-cpu-vulnerability/ and https://learn.microsoft.com/en-us/azure/virtual-machines/mitigate-spectre). Other Solution Notes ==================== n/a Dataflow ======== 48 clients were used to generate the workload plus 1 prime client.
Each workload client used a single virtual network adapter, through a single vnet connected to an Azure NetApp Files endpoint. Each client mounted one ANF large volume as an NFSv3 filesystem within the same Azure Region/availability zone.
Breakthrough mode enables six network endpoints per volume. The clients were round robin assigned to the available endpoints.
Other Notes =========== There is 1 mount per client, mounted at /mnt/eda on all 48 workload clients as listed in the CLIENT_MOUNTPOINTS line of the RC file. Example /etc/fstab entry:
10.0.10.4:/breakthrough-perf-test-lv /mnt/eda nfs hard,proto=tcp,vers=3,rsize=262144,wsize=262144,nconnect=8,nocto,noatime,actimeo=600 0 0
Other Report Notes ================== None =============================================================================== Generated on Mon Sep 14 16:28:53 2026 by SpecReport Copyright (C) 2016-2026 Standard Performance Evaluation Corporation