Single Client Performance - CIFS and NFS on Linux

A CentOS 6.2 virtual machine was used to evaluate NFS and CIFS performance of the NAS when accessed from a Linux client. In order to standardize the testing across multiple NAS units, the following parameters were used to mount the NFS and Samba shares:

mount -t nfs NAS_IP:/PATH_TO_NFS_SHARE /PATH_TO_LOCAL_MOUNT_FOLDER

mount -t cifs //NAS_IP/PATH_TO_SMB_SHARE /PATH_TO_LOCAL_MOUNT_FOLDER

Note that these areslightly different from what we used to run in our previous NAS reviews. We have also shifted from IOMeter to IOZone for evaluating performance under Linux. The following IOZone command was used to benchmark the shares:

iozone -aczR -g 2097152 -U /PATH_TO_LOCAL_CIFS_MOUNT -f /PATH_TO_LOCAL_CIFS_MOUNT/testfile -b <NAS_NAME>_CIFS_EXCEL_BIN.xls > <NAS_NAME>_CIFS_CSV.csv

IOZone provides benchmark numbers for a multitude of access scenarios with varying file sizes and record lengths. Some of these are very susceptible to caching effects on the client side. This is evident in some of the graphs in the gallery below.

Readers interested in the hard numbers can refer to the CSV program output here. These numbers will gain relevance as we benchmark more NAS units with similar configuration.

The NFS share was also benchmarked in a similar manner with the following command:

iozone -aczR -g 2097152 -U /nfs_test_mount/ -f /nfs_test_mount/testfile -b <NAS_NAME>_NFS_EXCEL_BIN.xls > <NAS_NAME>_NFS_CSV.csv

Some scenarios exhibit client caching effects, and these are evident in the gallery below.

The iozone CSV output can be found here for those interested in the exact numbers.

A summary of the bandwidth numbers for various tests averaged across all file and record sizes is provided in the table below. As noted previously, some of these numbers are skewed by caching effects. A reference to the actual CSV outputs linked above make the entries affected by this effect obvious.

WD My Cloud EX2 - Linux Client Performance (MBps)
IOZone Test CIFS NFS
Init Write 46 21
Re-Write 50 24
Read 23 92
Re-Read 23 94
Random Read 12 34
Random Write 30 18
Backward Read 13 28
Record Re-Write 30 243
Stride Read 21 63
File Write 49 22
File Re-Write 50 23
File Read 17 64
File Re-Read 17 67

 

Single Client Performance - CIFS and iSCSI on Windows Multi-Client Performance - CIFS
POST A COMMENT

20 Comments

View All Comments

  • lours - Tuesday, March 4, 2014 - link

    What is the default setup for the models that come with drives? What capacity are the drives? Reply
  • hlmcompany - Tuesday, March 4, 2014 - link

    As a whole, the unit comes in capacities of 0, 4, 6, and 8 TB. For the units that come with drives, they are the WD Red Drives, such as WD20EFRX, which are slower performing than the 7,200 rpm WD RE4 drives used in this review. Reply
  • ganeshts - Tuesday, March 4, 2014 - link

    Configurations with disks are ready to operate out of the box in RAID-1 configuration. So, a 4 TB model, say, will have 2 TB of usable space.

    As hlmcompany mentions, the WD Red drives are used.
    Reply
  • chubbypanda - Tuesday, March 4, 2014 - link

    Ganesh,

    Annoyingly enough, some WD previous models (My Book Studio II) would work only with WD Green drives. Looking at WD's list for EX2, situation is better in this case:

    http://www.wdc.com/en/products/products.aspx?id=11...
    Reply
  • Oyster - Tuesday, March 4, 2014 - link

    This review (including some of the other NAS reviews) seems lacking. What's the point of just showing the read/write numbers without giving us the benchmarks. Like why not show us the comparison to Synology and QNAP 2-bay NAS devices? I'd be more informative to see the performance of things like VM images running off these devices as well. I hope this feedback is accounted for in the future reviews. Reply
  • hlmcompany - Tuesday, March 4, 2014 - link

    Agreed. There might even be some relevancy to comparing the My Cloud EX2 with the previous generation model sold as the My Book Live. Reply
  • ganeshts - Tuesday, March 4, 2014 - link

    We have already taken care of that aspect in the multi-client tests. Each graph has a drop-down menu under it that allows readers to view the results from the evaluation of similar NAS units. For example, this review lists other 2-bay NAS units that we have evaluated in a similar manner before. The ioSafe N2 presented as a comparison in this review is the Synology DS213. Unfortunately, we haven't evaluated any 2-bay QNAP units.

    We also have NASPT single-client results for more NAS units. Currently working on a way to integrate a comparison with a drop-down similar to what we have done for the multi-client tests.

    As for VM images, can you clarify the exact use-case? I was under the impression that users mount iSCSI volumes on VMs, and we already present iSCSI performance numbers. (at least, that is what I do with some of my VMs).
    Reply
  • romrunning - Tuesday, March 4, 2014 - link

    To clarify, the comparisons with other 2-bay NAS are only on the "Multi-Client Performance - CIFS" page. It would have been helpful to have them on the other pages with performance graphs, like the Single-client Windows page. That would make it much easier to compare the different models/vendors. Reply
  • Oyster - Tuesday, March 4, 2014 - link

    Well, mounting the iSCSI volumes is good, but what I meant was to show us the actual comparisons of VM images running from different devices. So a use case would be to do a big Visual Studio compile on a VM over iSCSI on devices from different vendors. All these vendors claim virtualization compliance, so it'd be nice to put numbers against such claims. Data could include network throughput and CPU/RAM consumption, for e.g. This would also help us understand how good some of the Marvell processors are against Atom/x86. As it stands now, the graphs are too confusing...

    Also, as others have pointed out, it's a bit tedious to go clicking through every device from the dropdown and not at all intuitive. If I want to look at HD video playback throughput, I should be able to just look at the line graphs of the different vendors in one chart. Kind of like the video card comparison.
    Reply
  • creed3020 - Tuesday, March 4, 2014 - link

    I also have to agree that a comparison to other reviewed devices is lacking. This could be in the form of updates to the charts so that they show different units e.g. Video Cards FPS benchmarks, or done verbally within the conclusion or it's own section doing a breakdown of each test.

    These devices are not in Bench so we cannot compare the data there either.
    Reply

Log in

Don't have an account? Sign up now