[rescue] playing with the new E250 ... disk performance vs. old PeeCee

Sat Dec 16 16:28:07 CST 2006

OK well I have a better idea of what's in it now.  2x400MHz procs, 1GB 
RAM (i.e. half the banks are empty), 1x72GB 15,000 RPM IBM drive (i.e. 
no Sun disk label).

I have to check md5sums on my Solaris 10 DVD ISO or something because 
this box refuses to even try and boot off of the CD that I burned.  So I 
don't have any time with Solaris on this box yet.

I did however successfully get Ubuntu 6.06 LTS on it.  Some of the disk 
benchmarks are surprising.

I have a Pee Cee of similar vintage to the E250, an old Pentium III 
550MHz.  It's got a 300GB PATA drive on the on-board IDE controller. 
This Pentium box is not a server by any means.  Just a workstation I 
cobbled together from junk parts and is running as my home file server.

So running the same version of Linux on both machines, I ran "hdparm -T" 
and "hdparm -t" on the hard disk of each box.  I ran the commands 3x to 
get a more meaningful idea of what to expect.

Sun:
  Timing O_DIRECT cached reads:   732 MB in  2.01 seconds = 364.52 MB/sec
  Timing O_DIRECT cached reads:   744 MB in  2.00 seconds = 371.23 MB/sec
  Timing O_DIRECT cached reads:   728 MB in  2.00 seconds = 363.98 MB/sec
  Timing O_DIRECT disk reads:   52 MB in  3.03 seconds =  17.15 MB/sec
  Timing O_DIRECT disk reads:   56 MB in  3.10 seconds =  18.06 MB/sec
  Timing O_DIRECT disk reads:   50 MB in  3.04 seconds =  16.45 MB/sec

PeeCee:
  Timing cached reads:   540 MB in  2.01 seconds = 268.66 MB/sec
  Timing cached reads:   544 MB in  2.00 seconds = 272.00 MB/sec
  Timing cached reads:   536 MB in  2.00 seconds = 268.00 MB/sec
  Timing buffered disk reads:  138 MB in  3.02 seconds =  45.70 MB/sec
  Timing buffered disk reads:  148 MB in  3.04 seconds =  48.68 MB/sec
  Timing buffered disk reads:  146 MB in  3.03 seconds =  48.18 MB/sec

Do the low direct disk read numbers for the Sun box surprise anyone?  Am 
I hitting a limitation of the SCSI controller here?  I'm not too 
familiar with the E250 platform yet and I was thinking maybe it had a 
faster SCSI controller than that.  Looks more like it maxes out at 20MB/sec.

If I'm maxing out on the SCSI controller, is there a decent PCI SCSI 
card that is "correct" for these boxes that I can search for on eBay and 
plug into my hot swap backplane to achieve better performance?

I was honestly expecting better disk I/O performance out of the Sun. 
I'm a bit surprised.

Here is the "hdparm -I" output for the PeeCee drive (same command does 
not work on the Sun):

PeeCee:
ATA device, with non-removable media
         Model Number:       ST3300622A
         Serial Number:      3NF1QBR0
         Firmware Revision:  3.AAH
Standards:
         Supported: 7 6 5 4
         Likely used: 7
Configuration:
         Logical         max     current
         cylinders       16383   65535
         heads           16      1
         sectors/track   63      63
         --
         CHS current addressable sectors:    4128705
         LBA    user addressable sectors:  268435455
         LBA48  user addressable sectors:  586072368
         device size with M = 1024*1024:      286168 MBytes
         device size with M = 1000*1000:      300069 MBytes (300 GB)
Capabilities:
         LBA, IORDY(can be disabled)
         Queue depth: 1
         Standby timer values: spec'd by Standard, no device specific 
minimum
         R/W multiple sector transfer: Max = 16  Current = ?
         Recommended acoustic management value: 208, current value: 0
         DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 *udma4 udma5
              Cycle time: min=120ns recommended=120ns
         PIO: pio0 pio1 pio2 pio3 pio4
              Cycle time: no flow control=240ns  IORDY flow control=120ns
Commands/features:
         Enabled Supported:
            *    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
                 SET MAX security extension
            *    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
HW reset results:
         CBLID- above Vih
         Device num = 0 determined by the jumper
Checksum: correct

On the Sun, /proc/scsi/sym53c8xx/0 is thus:
Chip sym53c875, device id 0xf, revision id 0x14
At PCI address 0001:00:03.0, IRQ 5,7e0
Min. period factor 12, Wide SCSI BUS
Max. started commands 448, max. commands per LUN 64

It looks like a "Wide SCSI" controller if I believed this, but that 
would max it out at only 10MB/sec.  So I think I'm missing something here.

/proc/scsi/scsi on the Sun describes the hard disk as:
Host: scsi0 Channel: 00 Id: 00 Lun: 00
   Vendor: IBM-ESXS Model: DTN073C3UCDY10FN Rev: S27P
   Type:   Direct-Access                    ANSI SCSI revision: 03