NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
NFS Performance issues
I work for Netapp in the NFS support area.
I puchased this device almost a year ago and for the most part have been happy to have it just as a storage device.
Recently I tried to put a VM on the device. Mounted nfsv3, and default mount parameters. The mount is fine, however performance is absolutely terrible. I took a packet trace to see if it potentially could be network loss. There were a few packets that appear to be loss, but VERY little and they do not correspond with the latency that i'm seeing.
I am seeing upwards of 20 seconds for a single write call. I can send the trace out to you, but I dont see a way to attach it to this post. I downloaded gcc5.4 and attempted to untar. The untar process took 3min 54s to complete. The file was stored on the filer, so this would be a read and write. This was intentional to gauge performance.
I need assistance to see what I can do to get better performance out of this. I dont expect the world, but I do expect to be able to install updates on a windows vm in under 10 hours. (Local disk it is ~15 minutes). If there is anything / logs / traces / additional tools to run, please let me know.
8 Replies
Replies have been turned off for this discussion
Which firmware version are you running?
Did you follow the instructions here?: http://kb.netgear.com/app/answers/detail/a_id/28435As you can see in the screenshot bit-rot protection is disabled.
Have you also disabled snapshots on the share?
You may also find this article useful: http://kb.netgear.com/app/answers/detail/a_id/14765/~/how-to-enable-async-mode-in-readynas-os-6-for-improved-nfs-performanceApart from what mdgm mentioned (definitely worth a read), can you please let us know whether you have NFS async enabled or disabled?
Go to the settings page of the share --> Network Access --> NFS --> Advanced.
This can have a very significant performance impact - running with or without async.
Thanks
Hi all,
Thanks for the response. I looked over the KB's listed.
1) I am running 6.5.0 (1 minor version below current i believe)
2) ASYNC is not enabled (I'll play with this to see if this improves performance)
3) Bit-Rot Protection is not enabled.
4) Snapshots are enabled, set for 'smart' for management and are set to daily.
edit:
Additional outputs:
mount options
jtownsen@superjt:~$ mount | grep vm 192.168.1.79:/data/vm_store on /var/lib/kimchi/nfs_mount/vm_store type nfs (rw,relatime,vers=3,rsize=524288,wsize=524288,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=192.168.1.79,mountvers=3,mountport=59523,mountproto=udp,local_lock=none,addr=192.168.1.79)
top.
Tasks: 212 total, 1 running, 211 sleeping, 0 stopped, 0 zombie %Cpu(s): 0.2 us, 0.8 sy, 0.0 ni, 85.3 id, 13.7 wa, 0.0 hi, 0.1 si, 0.0 st KiB Mem: 3931280 total, 3245088 used, 686192 free, 312 buffers KiB Swap: 2094844 total, 0 used, 2094844 free, 1827016 cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 3656 root 20 0 0 0 0 D 1.7 0.0 0:23.35 nfsd 1504 root 20 0 0 0 0 S 1.3 0.0 49:32.66 md127_raid5 3655 root 20 0 0 0 0 S 0.7 0.0 0:12.35 nfsd 1443 root 0 -20 0 0 0 S 0.3 0.0 1:28.83 kworker/2:1H 2835 root 0 -20 2228m 17m 3120 S 0.3 0.5 16:10.14 vpnserver 2954 root 35 15 1830m 71m 6596 S 0.3 1.9 6:21.11 Plex Script Hos 2957 admin 20 0 2363m 137m 5188 S 0.3 3.6 14:43.74 sabnzbd 3543 admin 20 0 130m 7376 3084 S 0.3 0.2 0:02.97 apache2 14297 root 20 0 0 0 0 S 0.3 0.0 0:01.15 kworker/u8:5 14549 root 20 0 0 0 0 S 0.3 0.0 0:00.13 kworker/2:1 1 root 20 0 45720 5052 2488 S 0.0 0.1 0:06.58 systemdI enabled async on the volume.. Huge perf increase. 95 minutes to 4 minutes. I feel like this is indicating my disks are a problem.
root@superjt:/var/lib/kimchi/nfs_mount/vm_store# time tar xf gcc-5.3.0.tar.gz real 95m11.482s user 0m9.828s sys 0m48.948s root@superjt:/var/lib/kimchi/nfs_mount/vm_store# rm -rf gcc-5.3.0/ root@superjt:/var/lib/kimchi/nfs_mount/vm_store# time tar xf gcc-5.3.0.tar.gz real 4m7.290s user 0m8.152s sys 0m30.148s root@superjt:/var/lib/kimchi/nfs_mount/vm_store#
- I have 6x 3tb WD red drives
SuperJT wrote:
I have 6x 3tb WD red drivesThey seem unlikely to be the botteneck. With single redundancy they should be giving you 100 MB/sec for both reads and writes.
You could try nastester, that will take NFS out of the equation: http://www.808.dk/?code-csharp-nas-performance
Maybe download the logs and double-check the smart stats.
Did StephenB's post resolve your problem. If not, could you send me your logs (see the Sending Logs link in my sig)?
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
