| title | description | services | documentationcenter | author | manager | editor | ms.assetid | ms.service | ms.workload | ms.tgt_pltfrm | ms.devlang | ms.topic | ms.date | ms.author |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Azure NetApp Files performance benchmarks for Linux | Microsoft Docs |
Describes performance benchmarks Azure NetApp Files delivers for Linux. |
azure-netapp-files |
b-juche |
azure-netapp-files |
storage |
na |
na |
conceptual |
04/29/2020 |
b-juche |
This article describes performance benchmarks Azure NetApp Files delivers for Linux.
This section describes performance benchmarks of Linux workload throughput and workload IOPS.
The graph below represents a 64-kibibyte (KiB) sequential workload and a 1-TiB working set. It shows that a single Azure NetApp Files volume can handle between ~1,600 MiB/s pure sequential writes and ~4,500 MiB/s pure sequential reads.
The graph illustrates decreases in 10% at a time, from pure read to pure write. It demonstrates what you can expect when using varying read/write ratios (100%:0%, 90%:10%, 80%:20%, and so on).
The following graph represents a 4-kibibyte (KiB) random workload and a 1-TiB working set. The graph shows that an Azure NetApp Files volume can handle between ~130,000 pure random writes and ~460,000 pure random reads.
This graph illustrates decreases in 10% at a time, from pure read to pure write. It demonstrates what you can expect when using varying read/write ratios (100%:0%, 90%:10%, 80%:20%, and so on).
Linux 5.3 kernel enables single-client scale-out networking for NFS-nconnect. The graphs in this section show the validation testing results for the client-side mount option with NFSv3. The feature is available on SUSE (starting with SLES12SP4) and Ubuntu (starting with the 19.10 release). It's similar in concept to both SMB multichannel and Oracle Direct NFS.
The graphs compare the advantages of nconnect to a non-connected mounted volume. In the graphs, FIO generated the workload from a single D32s_v3 instance in the us-west2 Azure region.
The following graphs show sequential reads of ~3,500 MiB/s reads with nconnect, roughly 2.3X non-nconnect.
The following graphs show sequential writes. They indicate that nconnect has no noticeable benefit for sequential writes. 1,500 MiB/s is roughly both the sequential write volume upper limit and the D32s_v3 instance egress limit.
The following graphs show random reads of ~200,000 read IOPS with nconnect, roughly 3X non-nconnect.
The following graphs show random writes of ~135,000 write IOPS with nconnect, roughly 3X non-nconnect.
