October 27, 201015 yr I just finished putting together my new build. It's a Supermicro X7SPE-HF with 2GB of RAM. So far I have 3 WD20EARS (2TB) drives in the array with no parity. Pins 7-8 are jumpered on all drives. I mounted a 1TB drive that has a bunch of stuff to populate the array and am in the process of copying files over. Using mc I'm seeing on average about 30MB/s. Copying to the array across the network I am seeing about 40MB/s and reads are around 20MB/s. Am I just expecting too much out of this Atom or does this seem really low? Pics or it didn't happen right
October 27, 201015 yr I just finished putting together my new build. It's a Supermicro X7SPE-HF with 2GB of RAM. So far I have 3 WD20EARS (2TB) drives in the array with no parity. Pins 7-8 are jumpered on all drives. I mounted a 1TB drive that has a bunch of stuff to populate the array and am in the process of copying files over. Using mc I'm seeing on average about 30MB/s. Copying to the array across the network I am seeing about 40MB/s and reads are around 20MB/s. Am I just expecting too much out of this Atom or does this seem really low? Pics or it didn't happen right To copy a file from your PC there are two disks involved. The parity disk and the target disk. There are 4 I/O operations. The read of the parity disk, the read of the target disk, and the writing of the two disks. Using "mc" to copy a file from one disk to another requires one more I/O operation, that of reading the source disk. Other than your read speed, the others you mention are normal. It has absolutely nothing to do with the CPU. It has to do with the rotational speeds of the disks involved. See this post for details: http://lime-technology.com/forum/index.php?topic=4390.msg40684;topicseen#msg40684
October 27, 201015 yr Author To copy a file from your PC there are two disks involved. The parity disk and the target disk. There are 4 I/O operations. The read of the parity disk, the read of the target disk, and the writing of the two disks. Using "mc" to copy a file from one disk to another requires one more I/O operation, that of reading the source disk. Other than your read speed, the others you mention are normal. It has absolutely nothing to do with the CPU. It has to do with the rotational speeds of the disks involved. See this post for details: http://lime-technology.com/forum/index.php?topic=4390.msg40684;topicseen#msg40684 Shouldn't there should only be 2 IO ops (1 read and 1 write) because there is no parity drive enabled? I understand rotational speed plays a large part, but these drives should be able to sustain more than 30MB/s in write speed shouldn't they? I don't remember what preclear was reporting during the write phase, but it certainly wasn't this low. The 1TB source disk is a 7200 RPM drive so if the transfer is IO bound it is the EARS drives and not the source. I thought perhaps it has to do with the CIFS driver so I tried copying a file from /mnt/drive1 to /mnt/drive2 and am still only seeing 40MB/s copy rates. I might expect these speeds with the parity drive enabled but it is not hence my concern.
October 27, 201015 yr Author No worries, it's easy to do. I'll keep digging. I did try dd if=/dev/zero of=/mnt/drive2/file.tmp bs=1024 count=1000 to create a 1GB file and dd is reporting 80MB/s. Now I'm stumped
October 29, 201015 yr Author Update: I used dd to create a 1GB file dd if=/dev/urandom of=rndfile.tmp bs=1M count=1k 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 641.984 s, 1.7 MB/s and tried copying this file around with cp and using time to time the copy. I averaged 5 copies for the results. /mnt/disk2/Software/rndfile.tmp -> /mnt/disk1/Movies/rndfile.tmp = 69.1MB/s /mnt/user/Sofrware/rndfile.tmp -> /mnt/user/Movies/rndfile.tmp = 59.3MB/s Those numbers look a little more like what I would expect to see with no parity drive. For giggles I tried mc again and it was reporting copy speeds ~50-55MB/s. I don't know what has changed (other than reboots) but there it is. I don't think I have any problems but I snipped the info for one drive from the syslog and posted it and the whole syslog is posted as well. Oct 28 22:19:14 Nesoi kernel: ata1: SATA max UDMA/133 abar m2048@0xfebfb000 port 0xfebfb100 irq 33 Oct 28 22:19:14 Nesoi kernel: ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300) Oct 28 22:19:14 Nesoi kernel: ata1.00: ATA-8: WDC WD20EARS-00J2GB0, 80.00A80, max UDMA/133 Oct 28 22:19:14 Nesoi kernel: ata1.00: 3907029168 sectors, multi 0: LBA48 NCQ (depth 31/32), AA Oct 28 22:19:14 Nesoi kernel: ata1.00: configured for UDMA/133 Oct 28 22:19:14 Nesoi kernel: scsi 1:0:0:0: Direct-Access ATA WDC WD20EARS-00J 80.0 PQ: 0 ANSI: 5 Oct 28 22:19:14 Nesoi kernel: sd 1:0:0:0: [sda] 3907029168 512-byte logical blocks: (2.00 TB/1.81 TiB) Oct 28 22:19:14 Nesoi kernel: sd 1:0:0:0: [sda] Write Protect is off Oct 28 22:19:14 Nesoi kernel: sd 1:0:0:0: [sda] Mode Sense: 00 3a 00 00 Oct 28 22:19:14 Nesoi kernel: sd 1:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA Oct 28 22:19:14 Nesoi kernel: sda: sda1 Oct 28 22:19:14 Nesoi kernel: sd 1:0:0:0: [sda] Attached SCSI disk If you notice anything that looks out of place, let me know nesoi_syslog_2010-10-28.txt
Archived
This topic is now archived and is closed to further replies.