NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
Pro 6 reading performance
Hi there,
I'd like your opinion on a performance questions I have.
The NAS is a Pro 6 with two WDC RED 4TB.
If I send a big 30GB file to the NAS, I can see it goes through without slowing down. Check the graph below:
Now if I READ from it - and write on my HDDs, what often happens is this:
"that is the cache" you'd say. Ok. The point is that if I try again I have this
which takes quite a longer time to drop, and it does not drop till the awful levels of the previous graph.
I've done several tests and it seems that Windows at some point freaks out and decides to massively slow down the transfer. I can see my HDD slowly flashing when that happens. I can also "hear it" (I can hear my PC through the speakers and when the network is running I can hear a high pitched whistle which is steady when it's fast and intermittent when slows down).
I have an SSD drive on my PC, on that it takes more time to slow down, but it slows down nevertheless.
I know that officially I shouldn't reach 100MB/s but it seems to me the units are capable of that and it seems there is a software 'bug' somewhere, more likely to be Windows.
The ReadyNAS is basically factory default, I haven't enabled jumbo frames or touched anything else.
Any ideas?
Thank you!
I'd like your opinion on a performance questions I have.
The NAS is a Pro 6 with two WDC RED 4TB.
If I send a big 30GB file to the NAS, I can see it goes through without slowing down. Check the graph below:
Now if I READ from it - and write on my HDDs, what often happens is this:
"that is the cache" you'd say. Ok. The point is that if I try again I have this
which takes quite a longer time to drop, and it does not drop till the awful levels of the previous graph.
I've done several tests and it seems that Windows at some point freaks out and decides to massively slow down the transfer. I can see my HDD slowly flashing when that happens. I can also "hear it" (I can hear my PC through the speakers and when the network is running I can hear a high pitched whistle which is steady when it's fast and intermittent when slows down).
I have an SSD drive on my PC, on that it takes more time to slow down, but it slows down nevertheless.
I know that officially I shouldn't reach 100MB/s but it seems to me the units are capable of that and it seems there is a software 'bug' somewhere, more likely to be Windows.
The ReadyNAS is basically factory default, I haven't enabled jumbo frames or touched anything else.
Any ideas?
Thank you!
10 Replies
Replies have been turned off for this discussion
- What version of RAIDiator is running on the NAS and what version of Windows is running on the PC?
- Good question! I should have mentioned that earlier.
4.2.26 and Windows 7 64. I did some test before and I'm sure it was not like that. - If you download the logs (Status > Logs > Download all Logs), what do logs like initrd.log and network_settings.log look like?
Also what are the SMART+ stats for the disks? Is there a high load cycle count?
So you have just the two disks installed?
On your PC do you see any errors for transfers via the NIC when looking at the status for it? - initrd
[2014/05/11 21:28:26] Factory default initiated by button!
[2014/05/11 21:36:14] Selected X-RAID2 mode, 10GB snapshot, RAID level 5
[2014/05/11 21:37:12] Factory default initiated on RAIDiator 4.2.26.
network, I guess you are interested in the network I'm using
eth0 Link encap:Ethernet HWaddr xxxxxxxxxxxxxxxx
inet addr:xxx.xxx.xxx.xxx Bcast:xxx.xxx.xxx.xxx Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:72439698 errors:0 dropped:56 overruns:0 frame:0
TX packets:70852852 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:66872618114 (62.2 GiB) TX bytes:83483013130 (77.7 GiB)
Interrupt:16 Memory:ff6fc000-0
Settings for eth0:
Supported ports: [ TP ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Half 1000baseT/Full
Advertised auto-negotiation: Yes
Speed: 1000Mb/s
Duplex: Full
Port: Twisted Pair
PHYAD: 0
Transceiver: internal
Auto-negotiation: on
Supports Wake-on: g
Wake-on: d
Link detected: yes
How do I check NIC statistics on Windows??
my brand new switch does not report errors though.
Thanks! - So there is 56 dropped packets at the NAS end on the receiving end.
Not sure how you check it in Windows 7.
Under Status > Health > SMART+ what do the SMART+ stats for the disks look like (remove the disk serial number before posting, of course)? - Sorry I forgot to post the SMART details. I have already checked that the drives were not parking the heads too often, the value is set to 300s
I'll try to check windows errors on my NIC. Haven't found an easy way to do it yet!
I'm sure it was not happening before. The difference may be the new drives. I can try and stick in an old one and see if that happens anyway. I have also upgraded my NIC's driver, I'll test again later.
Thanks for your help!
Edit:
Ok, also found the way to read statistics under windows. it's
netstat -es (to reset statistics, disable and enable your NIC).
Here you go. During the transfer I monitored the NAS statistics. There was already a dropped package, and it stayed there. I believe dropped packages are not bad per se. Those are not errors, are just packages which did not apply/belong to that NIC. I also read somewhere that some Linux kernels/drivers (not a massive expert here) had a bug which was resulting in several dropped packages.
NAS statistics after a large transfereth0 Link encap:Ethernet HWaddr xxxxxxxxxxxxxx
inet addr:xxxxxxxxxxxxxx Bcast:xxxxxxxxxxxx Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:10095143 errors:0 dropped:1 overruns:0 frame:0
TX packets:18984628 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:693585858 (661.4 MiB) TX bytes:28151763483 (26.2 GiB)
Interrupt:16 Memory:ff6fc000-0
Windows statisticsnetstat -es -p IP
Interface Statistics
Received Sent
Bytes 657126268 2393232782
Unicast packets 75965830 40350488
Non-unicast packets 1048 20072
Discards 0 0
Errors 0 0
Unknown protocols 0
IPv4 Statistics
Packets Received = 19066880
Received Header Errors = 0
Received Address Errors = 0
Datagrams Forwarded = 0
Unknown Protocols Received = 0
Received Packets Discarded = 1756
Received Packets Delivered = 19070390
Output Requests = 10149103
Routing Discards = 1
Discarded Output Packets = 7111
Output Packet No Route = 68
Reassembly Required = 0
Reassembly Successful = 0
Reassembly Failures = 0
Datagrams Successfully Fragmented = 0
Datagrams Failing Fragmentation = 0
Fragments Created = 0
Windows is indeed showing lots of discarded packages, but I have no idea what that does mean. - The SMART stats look fine.
Does performance improve if more disks are installed (obviously would need to wait for the resync to complete and expansion to take place before you would know)? - Hi,
I haven't got more 4TB disks to try! :)
I am installing two 1TB right now (just those two drives) to rule the WDC out. - Ok, I have assessed that it's my RAID/Windows and not the NAS. When saving on the SSD it mostly goes flat out without a glitch - sometimes it slows down after a while, sometimes it doesn't.
I understand that writing on a RAID5 is slower than a normal drive, but I wouldn't have expected it to be so slow. It's an Intel software raid, so it's sort of expectable.
The upside is: I found a compatible 1GB DIMM module in a drawer! :) - Sorted.
My RAID did not have Write Back enabled! :)
Thanks for your help, appreciated.
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
