March 29, 201115 yr So after going through some issues setting up my unRaid, I finally have her up and running. I have it connected to the network via 1Gig ethernet. I mounted my Dlink 343 NAS directly to the unRaid server and using rsync to copy data over from the NAS to unRaid via a telnet console. I have the unRaid server setup via user shares... and i don't have a parity drive setup so this is unprotected... The data's been copying for 25 hrs now, and averaging ~15MB/s.. there are no errors in the syslog nor in the unRaid status page, and everything seems to be running properly but this seems a bit slow. I'm reading that on avg, people should be getting 25-40MB/s write speeds to unRaid ? I even set the frame size to 6k... (everything on my network has jumbo frames enabled set to 6k) It's a Biostar TA785G3HD motherboard so Realtek chipset, it's connected at 1000Mb/s... via a cat5e cable to the wall, which is connected via cat 5e to a Dlink 1024D switch which is then connected to the Dlink 343 NAS... which is also connected to the switch at 1000Mb/s I know copying from the Dlink 343 I can achieve read speeds around 24-25Mb/s... I can tell my the status that for now, it's writing to the Hitachi 7k2000 drive... so should be a fast drive... any suggestions ? at this rate, it's going to take over 3 days to copy my 4gigs of data over!
March 30, 201115 yr Unless you really know what you are doing, it is probably your attempt at enabling jumbo frames. They really will only help when there is a lot of latency. That is not typically the case on a home LAN. unRAID does not normally enable them. I'll bet you have all kinds of weirdness going with packet fragmentation. Post the output of ifconfig eth0
March 30, 201115 yr Author my home network seemed to have worked well... i tested transfer speeds with jumbo frame and without and was faster with jumbo frame. I also made sure that all the devices on the network supports jumbo frames at least up to 6k. Here's the results of the ifconfig eth0 command: eth0 Link encap:Ethernet HWaddr 00:30:67:60:b3:eb inet addr:192.168.0.114 Bcast:192.168.0.255 Mask:255.255.255.0 UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:6000 Metric:1 RX packets:252614371 errors:0 dropped:0 overruns:0 frame:0 TX packets:168486206 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:3576761662 (3.3 GiB) TX bytes:3566925474 (3.3 GiB) Interrupt:27 Base address:0x6000 Seems to be ok... no errors or overruns...
March 30, 201115 yr How are your reads? Have you tried disk performance tests, without involving the network? Just simple file copies from the command line, or maybe unmenu->User Scripts->Disk Speed Test.
March 30, 201115 yr Author i'm still in the middle of transferring 4gigs over so figured can't do any disk performance tests until that's done. I did try reading from the array and over the wired network was getting 55MB/s read.. which is double the speed of reading from the dlink343...
March 30, 201115 yr Attach a copy of your system log, perhaps there is something else going on. Also, post the output of ethtool eth0
March 31, 201115 yr Author Here are the results: root@Tower:~# ethtool eth0 Settings for eth0: Supported ports: [ TP MII ] 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: MII PHYAD: 0 Transceiver: internal Auto-negotiation: on Supports Wake-on: pumbg Wake-on: g Current message level: 0x00000033 (51) Link detected: yes and syslog attached. There are a few errors when mounting the disks, that is when I basically added the disks to the array i think and formatted them... syslog-2011-03-31.zip
April 1, 201115 yr Author added results of hdparm -tT command: root@Tower:/# hdparm -tT /dev/sdb /dev/sdb: Timing cached reads: 3872 MB in 2.00 seconds = 1936.67 MB/sec Timing buffered disk reads: 372 MB in 3.01 seconds = 123.73 MB/sec root@Tower:/# hdparm -tT /dev/sdc /dev/sdc: Timing cached reads: 3906 MB in 2.00 seconds = 1952.97 MB/sec Timing buffered disk reads: 372 MB in 3.01 seconds = 123.38 MB/sec root@Tower:/# hdparm -tT /dev/sdd /dev/sdd: Timing cached reads: 3760 MB in 2.00 seconds = 1880.64 MB/sec Timing buffered disk reads: 384 MB in 3.01 seconds = 127.50 MB/sec root@Tower:/# hdparm -tT /dev/sde /dev/sde: Timing cached reads: 3726 MB in 2.00 seconds = 1863.42 MB/sec Timing buffered disk reads: 326 MB in 3.02 seconds = 108.03 MB/sec root@Tower:/# hdparm -tT /dev/sdf /dev/sdf: Timing cached reads: 3726 MB in 2.00 seconds = 1863.00 MB/sec Timing buffered disk reads: 324 MB in 3.02 seconds = 107.35 MB/sec sdb = hitachi 7k2000 sdc = hitachi 5k3000 sdd = hitachi 7k2000 sde = seagate st3500320as sdf = seagate st3500320as Reads seem decent ?
April 1, 201115 yr Author you might have hit the nail on the head... Although I didn't check explicitly, I did try copying a large 13gig file from my desktop PC to unRaid and got around 75Mb/s transfer rate... which is a lot faster than I expected. I tried this several times, sometimes it was a bit slower down at 60Mb/s but 75Mb/s seemed the norm. (Caveat: This is without parity on writing to a Hitachi 7k2000 drive). I'm pretty happy with the results. It must be a speed issue between the Dlink 343 and unRaid, which I don't care about since I'm retiring the Dlink once the data is copied over. Solved!
April 1, 201115 yr you might have hit the nail on the head... I cheated and did a google search. I knew of the D-Link limitations during my previous search for my own Nas solution, and found a thread that trended CPU% with file transfers over SMB. ;-)
Archived
This topic is now archived and is closed to further replies.