Speed up lacie 5 Big Network 2

[Glaven Clattuck]

Hi to all,

Thanks to the active collaboration of the forum users, I restored my NAS.
however, i note that the file transfer performance are really very low.
This happened even before the crash. Then I ask myself and I ask all of you.
Is there a way to verify performance in the Lacie Transfer files?
My Mac and the NAS are connected to my LAN through a Router- Switch Asus RT-AC68U with port at 1GB.
On LAcie i use the LAN port 1.
I noticed using the command:

Code: Select all

mdadm --examine /dev/sd[abcde]2

that the Chunk Size is 64K.

Code: Select all

[root@LaCie-5big ~]# mdadm --examine /dev/sd[abcde]2
/dev/sda2:
          Magic : a92b4efc
        Version : 1.0
    Feature Map : 0x0
     Array UUID : ad468f58:53571e53:74558f1d:48545ec5
           Name : (none):4
  Creation Time : Tue Oct 28 21:06:48 2014
     Raid Level : raid5
   Raid Devices : 5

 Avail Dev Size : 7809984512 (3724.09 GiB 3998.71 GB)
     Array Size : 31239938048 (14896.36 GiB 15994.85 GB)
   Super Offset : 7809984768 sectors
          State : clean
    Device UUID : dc7d0611:5aa89423:0493d8ff:d0875d7d

    Update Time : Sat May 21 19:13:11 2016
       Checksum : f261c3d3 - correct
         Events : 698455

         Layout : left-symmetric
     [b][color=#FF0000]Chunk Size : 64K[/color][/b]

Can this parameter affect the data transfer?
I say this because I have noticed that other users had much higher values.
hoping for some suggestions
regards

[Jocko]

Hi Glaven Clattuck,

Indeed, chunk size may improve I/O access but with a 5 big2 is not obvious.
On a 5big2 you don't have a Raid control card but only logical raids. So in this case, CPU performances are the main limitation.
As CPU performances are not big on a 5big2, change chunk size value should have no effect.

Note: you can not change chunk size without losing your data.

[Glaven Clattuck]

Hi Glaven Clattuck,

Indeed, chunk size may improve I/O access but with a 5 big2 is not obvious.
On a 5big2 you don't have a Raid control card but only logical raids. So in this case, CPU performances are the main limitation.
As CPU performances are not big on a 5big2, change chunk size value should have no effect.

Note: you can not change chunk size without losing your data.

Thanks for your reply, specially for the last quote :-)

But my first question remain,

Is there a way to verify performance in the Lacie Transfer files?

We have some command in linux that can help on this way?

[fvdw]

What kind of LAN do you have? Gigabit or 100mbits/s or 10 mbits/s

You could check with the command "ifconfig" if your ethernet card interface has selected gigabit speed or 100 mbits/s or maybe even 10. Further it should be in full duplex mode.

Last if you transfer many small files transfer speed will be slow as well due to file handling.
I reach with fvdw-sl upto 50 MByte/s (400 mbits/s) for reading and writing up to 30 MB/s. With an experimental kernel reading goes up to values close to 90 MB/s

Lacie firmware shouldis also able to reach numbers of 50 MB/s for reading and 30 for writing on a gigabit LAN

[Glaven Clattuck]

What kind of LAN do you have? Gigabit or 100mbits/s or 10 mbits/s

You could check with the command "ifconfig" if your ethernet card interface has selected gigabit speed or 100 mbits/s or maybe even 10. Further it should be in full duplex mode.

Last if you transfer many small files transfer speed will be slow as well due to file handling.
I reach with fvdw-sl upto 50 MByte/s (400 mbits/s) for reading and writing up to 30 MB/s. With an experimental kernel reading goes up to values close to 90 MB/s

Lacie firmware shouldis also able to reach numbers of 50 MB/s for reading and 30 for writing on a gigabit LAN

My Lan is 1GB and the slowness is on transfer big file.
I try the Ifconfig (it's the same that the DOS command?)

[fvdw]

In dos it is ipconfig, and it is totally different
Ifconfig is used in lininux to didplay settins of the network interfaces

[Glaven Clattuck]

In dos it is ipconfig, and it is totally different
Ifconfig is used in lininux to didplay settins of the network interfaces

thanks for the explanation. This is my report, how do you think?

Code: Select all

[root@LaCie-5big ~]# ifconfig
bond0     Link encap:Ethernet  HWaddr 00:D0:4B:93:62:4F 
          inet addr:192.168.2.9  Bcast:192.168.2.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MASTER MULTICAST  MTU:1500  Metric:1
          RX packets:126139171 errors:0 dropped:0 overruns:0 frame:0
          TX packets:20482066 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:1409098017 (1.3 GiB)  TX bytes:1933264221 (1.7 GiB)

eth0      Link encap:Ethernet  HWaddr 00:D0:4B:93:62:4F 
          UP BROADCAST RUNNING SLAVE MULTICAST  MTU:1500  Metric:1
          RX packets:126139171 errors:0 dropped:0 overruns:0 frame:0
          TX packets:20482066 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:532
          RX bytes:1409098017 (1.3 GiB)  TX bytes:1933264221 (1.7 GiB)
          Interrupt:11

eth1      Link encap:Ethernet  HWaddr 00:D0:4B:93:62:4F 
          UP BROADCAST SLAVE MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:532
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)
          Interrupt:15

lo        Link encap:Local Loopback 
          inet addr:127.0.0.1  Mask:255.0.0.0
          UP LOOPBACK RUNNING  MTU:16436  Metric:1
          RX packets:233230 errors:0 dropped:0 overruns:0 frame:0
          TX packets:233230 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:23038941 (21.9 MiB)  TX bytes:23038941 (21.9 MiB)

[fvdw]

Looks normal, we need the commands

Code: Select all

ethtool eth0
ethtool eth1

To see details, hopefully ethtool is included in the lacie firmware

[Glaven Clattuck]

Looks normal, we need the commands

Code: Select all

ethtool eth0
ethtool eth1

To see details, hopefully ethtool is included in the lacie firmware

Yes. The ethtool command is supported!
There we are..

Code: Select all

[root@LaCie-5big ~]# ethtool eth0
Settings for eth0:
   Supported ports: [ ]
   Supported link modes:   
   Supports auto-negotiation: No
   Advertised link modes:  Not reported
   Advertised pause frame use: No
   Advertised auto-negotiation: No
   Speed: 1000Mb/s
   Duplex: Full
   Port: MII
   PHYAD: 8
   Transceiver: internal
   Auto-negotiation: off
   Link detected: yes


and

Code: Select all

[root@LaCie-5big ~]# ethtool eth1                                                                                                                     
Settings for eth1:
   Supported ports: [ ]
   Supported link modes:   
   Supports auto-negotiation: No
   Advertised link modes:  Not reported
   Advertised pause frame use: No
   Advertised auto-negotiation: No
   Speed: 10Mb/s
   Duplex: Half
   Port: MII
   PHYAD: 0
   Transceiver: internal
   Auto-negotiation: off
   Link detected: no


Appear that the Eth1, have a different conf...... :dry

[Jocko]

Hi Glaven Clattuck,

As you have enabled bonding feature that means 50% of transfer has a speed 10MB/s.

So I assume it is a wire issue. to valid this, swap the wires between eth0 and eth1. I expect speed transfer to be 10MB/s on eth0
In this case you will have to change it.

Otherwise disable bonding feature and use only eth0 port