February 25, 200719 yr I have tried quite a few things in the last 2 days ... I started with an AMD Abit KV7 board and two brand new WD 500GB sata300 drives. I installed UnRaid, configured it and it started the parity sync. With just the 2 drives I was getting 75,000KB/s speeds ... so it took 160ish minutes to complete. I thought this to be too slow... and thought the mobo was the problem. I swapped it with a AMD Asus A7N8X. I started it up and the parity sync dropped to 55,000KB/s. So a bit slower. So off I run to the local computer store and get the unRaid approved Sata300 contoller ... the Promise TX4 sata300. I installed it and I get the exact same results ... 55,000KB/s. I didnt want to get all hung up on parity sync ... I can live with that being slow. What I really care about is how fast I can get data on and off the server. I started up a large file transfer and all I am getting out of it is 3.55mB/s. When this was a windows pc I was getting much faster results. I am running a Gbit LAN My results from hdparm -I /dev/sdb is ... Tower login: root Linux 2.4.33. root@Tower:~# hdparm -I /dev/sdb /dev/sdb: ATA device, with non-removable media Model Number: WDC WD5000AAKS-00TMA0 Serial Number: WD-WMAPW1396640 Firmware Revision: 12.01C01 Standards: Supported: 7 6 5 4 Likely used: 7 Configuration: Logical max current cylinders 16383 16383 heads 16 16 sectors/track 63 63 -- CHS current addressable sectors: 16514064 LBA user addressable sectors: 268435455 LBA48 user addressable sectors: 976773168 device size with M = 1024*1024: 476940 MBytes device size with M = 1000*1000: 500107 MBytes (500 GB) Capabilities: LBA, IORDY(can be disabled) Queue depth: 32 Standby timer values: spec'd by Standard, with device specific minimum R/W multiple sector transfer: Max = 16 Current = 0 Recommended acoustic management value: 128, current value: 254 DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 udma5 *udma6 Cycle time: min=120ns recommended=120ns PIO: pio0 pio1 pio2 pio3 pio4 Cycle time: no flow control=120ns IORDY flow control=120ns Commands/features: Enabled Supported: * NOP cmd * READ BUFFER cmd * WRITE BUFFER cmd * Host Protected Area feature set * Look-ahead * Write cache * Power Management feature set Security Mode feature set * SMART feature set * FLUSH CACHE EXT command * Mandatory FLUSH CACHE command * Device Configuration Overlay feature set * 48-bit Address feature set Automatic Acoustic Management feature set SET MAX security extension * SET FEATURES subcommand required to spinup after power up Power-Up In Standby feature set * DOWNLOAD MICROCODE cmd * General Purpose Logging feature set * SMART self-test * SMART error logging Security: Master password revision code = 65534 supported not enabled not locked not frozen not expired: security count not supported: enhanced erase 128min for SECURITY ERASE UNIT. Checksum: correct root@Tower:~# and the rusults from a hdparm -i /dev/sdb is ... root@Tower:~# hdparm -i /dev/sdb /dev/sdb: HDIO_GET_IDENTITY failed: Inappropriate ioctl for device root@Tower:~# Any suggestions on how to get better speedds would be appreciated. Thanks!
February 25, 200719 yr Author Some other errors I am seeing... Tower login: root Linux 2.4.33. root@Tower:~# tail -f /var/log/syslog Feb 25 19:15:22 Tower smbd[1463]: [2007/02/25 19:15:22, 0] lib/util_sock.c:get_p eer_addr(1000) Feb 25 19:15:22 Tower smbd[1463]: getpeername failed. Error was Transport endp oint is not connected Feb 25 19:15:22 Tower smbd[1463]: [2007/02/25 19:15:22, 0] lib/util_sock.c:get_p eer_addr(1000) Feb 25 19:15:22 Tower smbd[1463]: getpeername failed. Error was Transport endp oint is not connected Feb 25 19:15:22 Tower smbd[1463]: [2007/02/25 19:15:22, 0] lib/util_sock.c:write _socket_data(430) Feb 25 19:15:22 Tower smbd[1463]: write_socket_data: write failure. Error = Co nnection reset by peer Feb 25 19:15:22 Tower smbd[1463]: [2007/02/25 19:15:22, 0] lib/util_sock.c:write _socket(455) Feb 25 19:15:22 Tower smbd[1463]: write_socket: Error writing 4 bytes to socke t 5: ERRNO = Connection reset by peer Feb 25 19:15:22 Tower smbd[1463]: [2007/02/25 19:15:22, 0] lib/util_sock.c:send_ smb(647) Feb 25 19:15:22 Tower smbd[1463]: Error writing 4 bytes to client. -1. (Connec tion reset by peer)
February 28, 200719 yr The rate reported for parity sync is the rate at which data is being written to the parity drive. A rate of 55,000KB/sec means 55MB/sec, which is pretty close to the raw transfer rate of your drive (assuming 7200RPM). You should see it start out fast, then as the parity sync progresses it will slow down due to the fact that drive transfer rate decreases as you reach the inner cylinders. Once you start adding more and more hard drives, the bottleneck becomes the speed of the PCI bus. Any single PCI bus can not transfer more than around 120MB/sec (assuming standard 33MH/32-bit PCI). Most motherboards actually have 2 PCI buses: one for the add-in cards, and one "internal" for the on-board disk controllers. Now you might think then we could read 240MB/sec, but actually the next bottleneck is the interface between the south bridge chip and the memory controller. This is typically around 200MB/sec (Intel, not sure what it is for AMD). As for your write transfer rate of 3.55MB/sec - yes this is not right, should be more like 15MB/sec. Are you sure your data path is entirely GigE? What program are you using to transfer, and how are you measuring?
March 2, 200719 yr Author You are right they are 7200RPM sata 300 drives ... good to know that I am peaking there. Are you sure your data path is entirely GigE? I was pretty sure ... both computers are reporting that they are at gig speeds and the switch is showing both are at gig as well. To make sure I hardwired the PC's together. I get the same slow speeds .. 3.5-4MB/s. I am using a Dlink ethernet card ... http://www.dlink.com/products/?sec=1&pid=284 I know this is not on your list... could it be causing problems?
March 3, 200719 yr Author Ok I have a little update.... The drive I was transfering the data from was bad ... in fact it went up in smoke! I got most of the data off so no big deal. I did some more network tests and here is what I found. I get about 9MB/s (off a good drive!) if I start up another transfer I get a total of 15MB/s ... so that looks good. My bottle neck looks like the PCI bus?? Is there anything I can do to help speed things up?
March 21, 200719 yr Ok I have a little update.... The drive I was transfering the data from was bad ... in fact it went up in smoke! I got most of the data off so no big deal. Don't you just love that smell... I did some more network tests and here is what I found. I get about 9MB/s (off a good drive!) if I start up another transfer I get a total of 15MB/s ... so that looks good. My bottle neck looks like the PCI bus?? Is there anything I can do to help speed things up? Try 4.0-beta1 & tell me what you get for read spead and write spead.
Archived
This topic is now archived and is closed to further replies.