Synology DS2015xs Review: An ARM-based 10G NAS
by Ganesh T S on February 27, 2015 8:20 AM EST- Posted in
- NAS
- Storage
- Arm
- 10G Ethernet
- Synology
- Enterprise
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. We chose IOZone as the benchmark for this case. In order to standardize the testing across multiple NAS units, we mount the CIFS and NFS shares during startup with the following /etc/fstab entries.
//<NAS_IP>/PATH_TO_SMB_SHARE /PATH_TO_LOCAL_MOUNT_FOLDER cifs rw,username=guest,password= 0 0
<NAS_IP>:/PATH_TO_NFS_SHARE /PATH_TO_LOCAL_MOUNT_FOLDER nfs rw,relatime,vers=3,rsize=32768,wsize=32768,namlen=255,hard,proto=tcp,timeo=600,retrans=2, sec=sys,mountaddr <NAS_IP>,mountvers=3,mountproto=udp,local_lock=none,addr=<NAS_IP> 0 0
The following IOZone command was used to benchmark the CIFS share:
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.
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
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.
Synology DS2015xs - Linux Client Performance (MBps) | ||
IOZone Test | CIFS | NFS |
Init Write | 86 | 82 |
Re-Write | 85 | 82 |
Read | 50 | 120 |
Re-Read | 50 | 122 |
Random Read | 33 | 70 |
Random Write | 80 | 82 |
Backward Read | 32 | 58 |
Record Re-Write | 56 | 1719* |
Stride Read | 45 | 117 |
File Write | 85 | 83 |
File Re-Write | 85 | 82 |
File Read | 35 | 95 |
File Re-Read | 36 | 97 |
*: Benchmark number skewed due to caching effect |
49 Comments
View All Comments
chrysrobyn - Friday, February 27, 2015 - link
Is there one of these COTS boxes that runs any flavor of ZFS?SirGCal - Friday, February 27, 2015 - link
They run Syn's own format...But I still don't understand why one would use RAID 5 only on an 8 drive setup. To me the point is all about data protection on site (most secure going off site) but that still screams for RAID 6 or RAIDZ2 at least for 8 drive configurations. And using SSDs for performance fine but if that was the requirement, there are M.2 drives out now doing 2M/sec transfers... These fall to storage which I want performance with 4, 6, 8 TB drives in double parity protection formats.
Kevin G - Friday, February 27, 2015 - link
I think you mean 2 GB/s transfers. Though the M.2 cards capable of doing so are currently OEM only with retail availability set for around May.Though I'll second your ideas about RAID6 or RAIDZ2: rebuild times can take days and that is a significant amount of time to be running without any redundancy with so many drives.
SirGCal - Friday, February 27, 2015 - link
Yes I did mean 2G, thanks for the corrections. It was early.JKJK - Monday, March 2, 2015 - link
My Areca 1882 ix-16 raid controller uses ~12 hours to rebuild a 15x4TB raid with WD RE4 drives. I'm quite dissappointed with the performance of most "prouser" nas boxes. Even enterprise qnaps can't compete with a decent areca controller.It's time some one built som real NAS boxes, not this crap we're seeing today.
JKJK - Monday, March 2, 2015 - link
Forgot to mention it's a Raid 6vol7ron - Friday, February 27, 2015 - link
From what I've read (not what I've seen), I can confirm that RAID-6 is the best option for large drives these days.If I recall correctly, during a rebuild after a drive failure (new drive added) there have been reports of bad reads from another "good" drive. This means that the parity drive is not deep enough to recover the lost data. Adding more redundancy, will permit you to have more failures and recover when an unexpected one appears.
I think the finding was also that as drives increase in size (more terabytes), the chance of errors and bad sectors on "good" drives increases significantly. So even if a drive hasn't failed, it's data is no longer captured and the benefit of the redundancy is lost.
Lesson learned: increase the parity depth and replace drives when experiencing bad sectors/reads, not just when drives "fail".
Romulous - Sunday, March 1, 2015 - link
Another benefit of RAID 6 besides 2 drives being able to die, is the prevention of bit rot. In Raid 5, if i have a corrupt block, and one block of parity data, it wont know which one is correct. However since RAID 6 has 2 parity blocks for the same data block, its got a better chance if figuring it out.802.11at - Friday, February 27, 2015 - link
RAID5 is evil. RAID10 is where it's at. ;-)seanleeforever - Friday, February 27, 2015 - link
802.11at:cannot tell whether you are serious or not. but
RAID 10 can survive a single disk failure, RAID 6 can survive a failure of two member disks. personally i would NEVER use raid 10 because your chance of losing data is much greater than any raid that doesn't involve 0 (RAID 0 was a afterthought, it was never intended, thus called 0).
RAID 6 or RAID DP are the only ones used in datacenter for EMC or Netapp.