Page 1 of 2 1 2 LastLast
Results 1 to 20 of 33

Thread: SCSI/SAS Controllers

  1. #1
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb SCSI/SAS Controllers

    STORAGE GUIDE

    To make sure everything's in tip-top shape, I would first disconnect the SCSI sub-system, then thoroughly repair your startup drive and any other internal drive with a repair permissions, First Aid from Disk Utility, and Diskwarrior.

    Follow Kaye's advice to break the UL4D RAID into individual SoftRAID-initialized drives and test each. This way, you can rule out problems with the software and focus on the sub-system's hardware.

    How to Install PCI/AGP Cards
    Mac OS X Available Firmware


    Storage Resources

    ATTO Update
    Active Termination
    Storage Reference
    Ultimate Boot Drive
    Tagged Command Queueing
    Initio Miles HBAs
    Adaptec
    Resetting NVRAM
    Resizing Partitions
    Low Level Format Explained
    Installing A Drive
    Guide to Storage
    RAID FAQ
    Build Your Own RAID
    RAID G5
    NCQ SATA drives and controllers
    External SATA, SAS Interface
    S.M.A.R.T. FAQ Attributes, Documentation
    Storage FAQ
    www.storagereview.com leaderboard and performance DB.
    http://www.storagereview.com/comparison.html
    Reference Guide
    Seagate Barracuda IV RAID
    SATA or SCSI on Desktop
    TCQ, RAID, SCSI and SATA
    SATA vs SCSI
    G5 PCI Compatibility
    Initio MilesU2W
    SATA - eSATA
    SATA FAQ
    FirmTek SeriTek/1S2
    Creative Cow HUGE Systems
    jumper settings for SCSI and IDE (PDF) manual

    Seagate FAQ

    Seagate Serial ATA FAQ
    Seagate Native Command Queueing
    How to use Cable Select
    Using Cable Select, a single drive will perform as master (ID 0) even when placed on the gray middle connector.
    --------

    Western Digital WD2500JB
    X-bit Labs Benchmark SATA
    Seagate Savvio 2.5-inch drive SAS/SATA
    Techreport Maxtor's DM10 NCQ
    SATA RAID Performance - why RAID doesn't boost load times
    Barefeats: Ultimate Boot Drive
    SATA vs PATA vs FW800
    LaCie FW800

    Addonics IDE to Serial ATA converter

    Hitachi K7250 SATA
    Hitachi "Vancouver" 180GXP-T7K250
    Best of Class: 7K250/7K400
    Hitachi Defective Drive Sounds

    RAPTOR

    Raptor, 7K250, Cheetah
    SR Review WD Raptor 73GB (2004)
    Raptor compared to 10K/15K SCSI
    Raptor WD740GD vs Atlas 10K V


    ATTO

    1. Make sure you are using the latest ATTO Configuration Tool
    2. Update the UL4D driver
    3. Flash the UL4D card to the latest available firmware ("flash file")
    4. Make sure you are running the latest version of SoftRAID

    Will ATTO ExpressPCI host adapters with a standard 33 MHz or 66 MHz PCI interface (not PCI-X) work in a G5 system?

    Certainly, with the exception of older adapters designed for 5-Volt-only PCI buss. ATTO has released updated drivers to support our Ultra3 and FC 3321 host adapters. These drivers are available on our driver download site.

    How does using a PCI adapter in a PCI-X slot affect performance?

    The three G5 PCI-X slots are configured as two independent banks. One bank (comprising one slot) operates at 133 MHz, and the remaining bank (comprising two slots) operates at 100 MHz. The standard UL3S/UL3D host adapter is appropriate for the entry-level G5 because both support 33 MHz PCI, which results in a good match.

    Each of our currently available PCI Fibre Channel adapters supports 66 MHz PCI operation.
    For best performance, use these with the high-end G5 systems.
    However, they will work in the entry-level G5 at 33 MHz speeds.
    http://www.attotech.com/troublemac.html#g5

    ATTO Software
    ATTO Mac Support
    ATTO Utility Software
    ATTO OS X Config Tool
    ATTO Drivers
    Last edited by TZ; 12-01-2005 at 01:51 PM.

  2. #2
    Join Date
    Nov 2004
    Location
    Germany
    Posts
    2,352

    Default

    I was using Adaptec SCSI cards in my Macs since 1990.

    The Adaptec cards 39160, 29160, 2930, 2906 are working more or less in OS X (some users have problems that connected devices are not showing up).
    I personally never had any problems with Adaptec cards since 10.3.

    Some SCSI cards don't work with 10.3.3. With Panther version 10.3.4 and higher there should be no problem.

    I have tested the following cards:

    1) ATTO PCI Express PSC

    2) ATTO UL3D

    3) Adaptec 2906

    4) Adaptec 2930CU (Rev C Rom version 4.0)

    5) Advansys 3940UA (in 10.3.4 the X Boot and the 9 and X non boot ROM works)

    6) Advansys 3940UW (No further driver dev. EOL)

    7) Initio INI9100UW "Miles" Rev. F

    8) Apple ATTO PCI Express PSC ROM V1.2

    Conclusions:

    1) ASP 10.3.4 shows all informations, works right out of the Box great UW card, works in 9, 10.2.8, 10.3.3, 10.3.4, 10.3.5

    2) ASP 10.3.4 shows all informations, works right out of the Box great dual channel U160 card, works in 9, 10.2.8, 10.3.3, 10.3.4, 10.3.5

    3) ASP 10.3.4 don't shows SCSI Bus informations, in 9, 10.2.8 and 10.3.4, 10.3.5 it works. No go in 10.3.3.

    4) ASP 10.3.4 don't shows SCSI Bus informations, in 9, 10.2.8 and 10.3.4, 10.3.5 it works. No go in 10.3.3.

    5) ASP 10.3.4 shows all informations, don't know where to buy it, Advansys was bought from Initio, works great in 9, 10.2.8, 10.3.3, 10.3.4, 10.3.5 only downside, if you want to switch between 9 and X you can not boot

    6) ASP 10.3.4 shows all informations, works great in 9 and 10.2.8 only downside, if you want to switch between 9 and X you can not boot product is EOL no further development on Panther 10.3

    7) ASP 10.3.4 shows all informations, works great in 9, 10.2.8, 10.3.3, 10.3.4, 10.3.5 only downside, if you want to switch (and boot from a HD connected to the card) between 9 and X you have to do a firmware udpate (X to 9, 9 to X firmware) each time you switch.

    8) ASP 10.3.4 shows all informations, works right out of the Box great UW card, works in 9, 10.2.8, 10.3.3, 10.3.4, 10.3.5 (used in G3 Servers)

    All cards tested in my MDD.

    BTW just take a look at this site: http://www.dantz.com/index.php3?SCRE...KBASE&id=27381

    from Dantz:

    Note: Dantz does not support any Adaptec SCSI adapters for use in any Macintosh computers running Mac OS X 10.3 Panther. Adaptec is no longer developing SCSI drivers for the Mac OS.

    Adaptec has issued a driver compatibility statement at:

    http://adaptec-tic.adaptec.com/cgi-b...?p_faqid=11052

    Best Regards

    Nicolas

  3. #3
    Join Date
    Jan 2002
    Location
    NW Montana
    Posts
    8,197

    Default

    Good question. I hope so. I have a Miles U2W with 2 - 10K Cheetahs and would like to put them into a first generation AGP G4 Sawtooth - if I part out part my 7500 - G3/400. Actually leaving my G3/400 and getting newer faster ATA/SATA drives would be better faster.

    Randy

  4. #4
    Join Date
    Sep 2004
    Location
    Compton CA
    Posts
    387

    Default Sorry to tell you but I am not aware of a way.

    http://www.macgurus.com/forums/showthread.php?t=19305
    This thread here was the same problem end result was moving away from Initio.
    If it wasn't for you I wouldn't be me

  5. #5
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    Nicolas,

    You are correct, no work in Beige G3 and OSX http://initio.com/support/download.htm#mac k

  6. #6
    Join Date
    Aug 2001
    Location
    Grangeville, ID USA
    Posts
    9,109

    Default

    From what I hear at Initio, unofficially of course, there is a patch kernel that can be used to get it to work. That is available at either xlr8yourmac or macfixit.

    I have never had a Beige, so personal knowledge is not mine, in this case. I can only wish you luck after passing the info on.

    Rick
    molṑn labe'
    "I am a mortal enemy to arbitrary government and unlimited power. I am naturally very jealous for the rights and liberties of my country, and the least encroachment of those invaluable privileges is apt to make my blood boil."
--Ben Franklin

  7. #7
    Join Date
    Nov 2004
    Location
    Germany
    Posts
    2,352

    Default

    Thank you all for your reply.

    Ricks, I was looking for the patch but I don't find it. I've found a Beta OSX patch for the Beige G3's.

    Mac Japan, the Miles needs to be flashed from OS9 to work in Sytem 10. The App. you mentioned, is needed to get it back to work in OS9.
    Each time you switch between 9 and 10 you have to flash the firmware.
    I tried it with my G4 MDD.

    This Beige (not mine it, a friends Mac) is used as a AFP fileserver (3 Clients) running 10.2.6 (single user version not server).
    In this Mac was an Apple Adaptec 2940 installed wich died 2 days ago.

    The card was connected to a RAID Tower case (6x 160GB IBM's 64MB cache RAID5 with 5 drives 1 hot spare, one system 4GB and one data partition 7xxGB)

    To get things back to work, I just installed an Advansys 3940AU and a 9GB HD (with an OS9 volume and and OS10 volume) connected to the onboard SCSI (if you have to reset the Beiges PRAM the Mac is only able to boot this OS9 drive)

    We tried an ATTO UL4S wich shows up in 10 but refuses to boot the system.

    Best Regards

    Nicolas

  8. #8
    Join Date
    Nov 2004
    Location
    Germany
    Posts
    2,352

    Default

    Hello Randy,

    the Miles U2W should be work without a problem in the B&W.
    Maybe you have to flash the firmware as you have to on the Miles UW.
    The problem that the card is not working is only Beige and Miles UW related.

    Best Regards
    Nicolas

  9. #9
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    Nicolas,

    Now it is beginning to dawn on me that you have a Miles UW, not a Miles U2W.

    Did you check your Miles UW that it is a revision F board and not a revision E?

    Looking here and scroll further http://initio.com/support/download.htm#mac the revision E will never work. The reason is that the revision E board firmware chip that holds the OS firmware update is too small to hold the OS X firmware. Only the revision F board firmware chip is large enough for the OS X firmware update. k

  10. #10
    Join Date
    Nov 2004
    Location
    Germany
    Posts
    2,352

    Default

    Hello Kaye,

    it is the Rev. F version of the card.
    But it is a known problem, I think OpenFirmware related.

    The Initio Miles UW card is not even working with Linux.
    When I tried to install Linux (Mandrake 8.1 and 9.1, YDL 3.xx) the system hangs in the step "partitioning your drives" with the card removed the installation went fine.

    I will try to get a used ATTO UL2 card for this friends Beige.

    To proof, what I have told you, I have tried a Miles U2W in a B&W 400MHz Rev.2 yesterday and it works great.
    Throughput is great too (72MB/s to 76MB/s with only one new 73GB 15k Cheetah)
    This B&W is System 10.2 only. I have flashed the card in a Beige before I installed it in the B&W.

    Best Regards
    Nicolas

  11. #11
    Join Date
    Aug 2000
    Location
    Concord, CA
    Posts
    7,056

    Default

    Well that's a bummer about the MilesUW not working in the beige. I have a couple of rev.F cards in my parts bin. At the time I used the MilesUW, it was the only SCSI card with internal and external 68-pin connectors and one internal 50-pin connector. Eventually that outlived its usefulness for me. Yes the MilesU2W is a great card as are the ATTO cards. k

  12. #12
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb Initio MilesUW/U2W

    MacGurus SCSI Initio

    About Miles UW:

    Now it is beginning to dawn on me that you have a Miles UW, not a Miles U2W.
    Did you check your Miles UW that it is a revision F board and not a revision E?

    This thread here was the same problem end result was moving away from Initio.
    http://www.macgurus.com/forums/showthread.php?t=19305

    Looking here and scroll further http://initio.com/support/download.htm#mac the revision E will never work. The reason is that the revision E board firmware chip that holds the OS firmware update is too small to hold the OS X firmware. Only the revision F board firmware chip is large enough for the OS X firmware update. k

    it is the Rev. F version of the card. But it is a known problem, I think OpenFirmware related.

    The Initio Miles UW card is not even working with Linux.
    When I tried to install Linux (Mandrake 8.1 and 9.1, YDL 3.xx) the system hangs in the step "partitioning your drives" with the card removed the installation went fine.

    I will try to get a used ATTO UL2 card for this friends Beige.

    To proof, what I have told you, I have tried a Miles U2W in a B&W 400MHz Rev.2 yesterday and it works great. Throughput is great too (72MB/s to 76MB/s with only one new 73GB 15k Cheetah) This B&W is System 10.2 only. I have flashed the card in a Beige before I installed it in the B&W.

    - Well that's a bummer about the MilesUW not working in the beige.

    I have a couple of rev.F cards in my parts bin. At the time I used the MilesUW, it was the only SCSI card with internal and external 68-pin connectors and one internal 50-pin connector. Eventually that outlived its usefulness for me. Yes the MilesU2W is a great card as are the ATTO cards.

    the Miles U2W should be work without a problem in the B&W.
    Maybe you have to flash the firmware as you have to on the Miles UW.
    The problem that the card is not working is only Beige and Miles UW related.

    the Miles needs to be flashed from OS9 to work in Sytem 10.
    The App. you mentioned, is needed to get it back to work in OS9.
    Each time you switch between 9 and 10 you have to flash the firmware.
    I tried it with my G4 MDD.

    This Beige (not mine it, a friends Mac) is used as a AFP fileserver (3 Clients) running 10.2.6 (single user version not server). The card was connected to a RAID Tower case (6x 160GB IBM's 64MB cache RAID5 with 5 drives 1 hot spare, one system 4GB and one data partition 7xxGB)

    To get things back to work, I just installed an Advansys 3940AU and a 9GB HD (with an OS9 volume and and OS10 volume) connected to the onboard SCSI (if you have to reset the Beiges PRAM the Mac is only able to boot this OS9 drive)

    We tried an ATTO UL4S wich shows up in 10 but refuses to boot the system.

    From what I hear at Initio, unofficially of course, there is a patch kernel that can be used to get it to work. That is available at either xlr8yourmac or macfixit.

    I have never had a Beige, so personal knowledge is not mine, in this case. I can only wish you luck after passing the info on. - Rick

    "I use an Initio Miles Bluenote card in my Quicksilver dual 1 GHz G4 to run my Microtek Scanmaker 5. With 'Tiger' it no longer works although the card is visible to the system profiler. Initio are working on a
    Please Note:
    Beginning in late March 2003 all Initio SCSI cards come with PC and OS 10 firmware on them. If you are running OS 9 with one of these cards you will have to flash the firmware on the card so it will work under OS 9. You can download the OS 9 firmware here

    Apple G5 Users Please Note:
    Apple has three G5 models, the 1.6 GHz, 1.8 GHz and 2GHz.
    In our testing we have found that only our MilesU2W will work in these machines, the Bluenote and Miles are not compatible with the PCI-X slots used in the new G5.

    How To Update the Firmware Mac SCSI Adapters

    1. Download the current versions of the SmartSCSI and SIM file for your respective adapter
    2. Next unStuff both files to the desktop
    3. Run the SmartSCSI application and select Setup/1st card configuration. The title bar on the next screen will list the version of the currently installed firmware
    4. Click on the button called Update Firmware near the bottom of the screen
    5. After about 10 seconds you will get a message that the update was successful
    6. Now click on the Load Defaults button and close all windows. Reboot the machine

    Sometimes after you update the firmware, the system will refuse to boot up correctly.

    If you own an old world machine (7500 to Beige G3) you will want to zap the PRAM to correct this. Click here to find out how!

    If you own a Blue & White G3 or G4 you will want to boot into Open Firmware and reset the hardware configuration.

    Resetting PRAM and NVRAM on Power Macintosh and iMac

    1. Shut down the computer.
    2. Locate the following keys on your keyboard: Command, Option, P, and R.
    You will need to hold these keys down simultaneously in step 3.
    3. Press the Power button, then hold down the Command-Option-P-R key
    combination. You must press this key combination before the gray screen appears.
    4. Hold the keys down until the Macintosh restarts itself twice.
    5. Release the keys after the Macintosh restarts twice.
    6. The parameter RAM and the non-volatile RAM in your computer is reset to the default values. The clock setting are not changed.

    Zapping PRAM


    The following will force it to see the new hardware.

    1) Shut down the computer.
    2) power up in open firmware. (hold "cmd-alt-O-F") You'll see a blank screen with blinking cursor.
    3) type "reset-nvram" and hit return.
    4) type "set-defaults" and hit return.
    5) type "reset-all" and hit return.

    Your cpu should now reboot and it should list what it found in its pci slots.
    Instructions here!

    If it does not work, make sure that your SCSI card is properly seated in the PCI slot, verify your cabling and termination as well as SCSI ID's of the connected devices.
    http://www.initio.com/support/index-download.htm
    Initio's firmware ReadMe states: "Apple made a change
    Initio MilesU2W and Tiger
    http://initio.com/support/download.htm#mac
    Miles U2W "DA Fix" 3.04 beta (2001)

    http://www.xlr8yourmac.com/SCSI/Ultra2SCSI/
    http://www.xlr8yourmac.com/archives/may01/050301.html
    http://www.xlr8yourmac.com/SCSI/darinsU2.html
    http://www.xlr8yourmac.com/archives/nov99/112099.html
    http://www.xlr8yourmac.com/scsi_.html

    http://www.macintouch.com/bg3.html
    http://www.macintouch.com/tiger40.html
    http://www.macintouch.com/g5reader08.html
    http://www.macintouch.com/backup02.html
    Last edited by TZ; 12-01-2005 at 01:56 PM.

  13. #13
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb Serial Attached SCSI

    Supports both Serial Attached SCSI and Serial ATA hard disk drives.

    Tom's Hardware: SAS Storage

    SAS hard drives will be used in enterprise environments (e.g., server storage, high-end workstations), where multiple users are accessing a single system. These mission-critical installations will require the highest levels of system performance for applications such as online transac-tion processing, data analysis and storage virtualization.

    Compatibility between the new serial interfaces benefits system builders and end users alike. The common serial interconnect will enable system builders to deploy systems that share common backplanes, connectors and cabling. This greatly simplifies the process of changing the
    "mix" of drives within an enclosure, enabling users to easily replace a SATA drive with a SAS drive if their needs change.

    End users will have additional flexibility in buying systems that are optimally configured to address their target application. The ability to replace a SATA drive with a SAS drive also eliminates the need to replace entire systems when additional performance/reliability is required.
    These factors combine to substantially reduce the total cost of ownership for desktop and enterprise storage.
    Hitachi SAS pdf
    Dual-Port SAS Drives Boon to IT pdf
    Currently, a 15,000-RPM disk drive will sustain data rates up to 75 megabytes/sec. At these sustained data rates, two disk drives will saturate a Serial ATA 1.5 gigabits/second bus.

    A Serial Attached SCSI controller with wide ports will deliver much higher bandwidth than Serial ATA. The Serial Attached SCSI architecture is designed to support over 16,000 physical links. The ability to aggregate bandwidth through the use of wide ports will provide the performance scalability required by next generation servers and storage systems. Performance
    White Papers
    Understanding SAS

    Q. What's the difference between Serial Attached SCSI and Serial ATA?

    Serial ATA is cost-optimized for the desktop as a direct replacement for EIDE. Adaptec expects Serial ATA hard disk drive performance and reliability to match that of today's desktop drives. Pricing for enterprise-class Serial Attached SCSI drives is expected to be no higher than today's SCSI and Fibre Channel drives.

    Q. Will servers use Serial ATA or Serial Attached SCSI?

    SCSI still dominates the mainstream server storage market. Parallel SCSI and Serial Attached SCSI will continue to provide the most value for enterprise applications where reliability, availability and scalability are key requirements.

    Over the past two years, EIDE penetration into the cost-sensitive sub-entry server market has increased, a trend expected to continue with Serial ATA. Serial ATA offers enhancements relative to EIDE, but remains inappropriate for enterprise environments.

    Q. Does Serial Attached SCSI affect the market for Ultra640 SCSI?

    Yes. Adaptec expects Serial Attached SCSI to replace Parallel SCSI over time. Whether or not customers deploy Ultra640 depends on the timing of Serial Attached SCSI product availability. FAQ
    Questions about Serial Attached SCSI

    Adaptec SAS Education Center

    Seagate admits it is aiming its strategy at firms that want to fiddle with the technology before they deploy it. Were positioning this bundle as a starter kit for customers to experiment with SAS, says Franco Castaldini, Seagates senior product marketing manager. The intention is to show how easy it is to upgrade from SCSI to SAS, he adds. Seagate Prepares SAS Cheetah 15K.4

    Serial Attached SCSI hard drives are expected in early 2006 from Fujitsu, Maxtor, Seagate, and Hitachi. The Maxtor Atlas 10K V is available now, and sports the same performance as SCSI only, SAS interface connect, and 16MB cache, and SCSI reliability.
    Last edited by TZ; 04-13-2006 at 07:10 AM.

  14. #14
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb SAS Hard Drives

    Hitachi Ultrastar 15K147
    Capacity 147GB / 73GB / 36 GB
    Configuration flexibility

    Interfaces Ultra320 SCSI (68 et 80 pin), 4GFC, 2GFC and Serial Attached SCSI (SAS)
    Ease of integration and configuration lexibility

    Data buffer (MB)2 16
    Rotational speed (RPM) 15,000
    Latency average (ms) 2
    Media transfer rate (max. Mbits/sec) 1129
    Interface transfer rate (MB/sec, max) 320 (SCSI) 300 (Serial)
    Sustained data rate (MB/sec.max) 93.3

    Seek time (read, typical)3
    Average (ms) 3.7 / 3.6 / 3.3
    Track to track (ms) 0.6
    Full stroke (ms) 6.7

    Reliability
    Error rate (non-recoverable) 10 in 10E16
    Start/stops (at 40° C) 50,000

    Acoustic
    Idle (Typical (Bels)) 3.5
    Operating (Typical (Bels)) 4.6

    Hitachi Ultrastar 15K147 36GB U320-Serial Attached SCSI (SAS) 15,000RPM Hard Drive w/16MB Buffer $223

    Maxtor Atlas 10K V 73GB 16MB Cache - $224 www.maczone.com

  15. #15
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb Adaptec

    We covered some issues with SCSI devices in our Mac OS X 10.4.0 special report, and identified a workaround, successful for many users, which involves removing the following files from the /System/Library/Extensions folder:

    * Adaptec 290X-2930.kext
    * Adaptec 29160x.kext
    * Adaptec 39160x.kext

    but leaving the file

    * Adaptec 78XXSCSI.kext

    in place, then restarting.

  16. #16
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb SATA Features: SSC, NCQ

    TCQ, NCQ Feature

    From www.StorageReview.com -
    Ever since Western Digital's Raptor WD740GD was announced, oh, about 10 months ago, enthusiasts and IT professionals everywhere have been patiently (or otherwise) waiting for concrete results that demonstrate the potential benefits of the the drive's tagged command queuing (TCQ).
    We've been embroiled in hundreds of hours of testing using several different TCQ-enabled controllers in conjunction with arrays of 1 to 4 WD740GD drives. Unfortunately, many hours of tests remain.

    However, we're eager to present some initial results. Stay tuned for the first of a three-part series that will examine what benefits TCQ confers in our relatively modest third-generation testbed in single-user and multi-user setups!
    And you thought it was easy?

    A good discussion on technology and differences, and how ATA gets around inability to do SCSI-style reordering of I/O and isn't really the same animal when it comes to tagged command queueing that we keep hearing about. From the Storage Review Computing Forum
    Starting next month (June 2004), Seagate's Barracuda SATA HD (200GB only) will integrate a new function/feature named: Native Command Queuing. It allows instruction to be reorganized while the are queuing in order to optimize their executions. Based on a test, a NCQ HD @ 7200rpm could challenge a Raptor 74GB, and that's quite a result. This function/feature could be really useful for MacOSX-based computers since the system and multitask processes are performing numerous disk access request.

    www.hardmac.com
    A good discussion on technology and differences, and how ATA gets around inability to do SCSI-style reordering of I/O and isn't really the same animal when it comes to tagged command queueing that we keep hearing about.
    Any form of tagged command queuing, by nature, requires firmware support in both the controller and the hard drive. It is also, necessarily part of the interface. An interface isn't simply a set of wires. It is a set of specifications that defines and enables the movement of data. If you want to modify the manner in which data is sent then you must account for it in the interface specification. Tagged Command Queuing is never a driver feature. Enabling controller support may require driver support in some implementations, but such implementations are only turning on or off a firmware feature, and since hard drives don't have drivers, in the end, you always need firmware support.
    I don't think a normal controller would bother implementing 32 registers when it only needs 1 normally. Normal ATA read and write commands must be completed before another command can be issued so there is never a need for maintaining Tags or DMA data on a non-TCQ-enabled controller. Any controller that implements these elements obviously possesses specific hardware support for it even if a driver is needed to enable it.
    From the Storage Review Computing Forum
    A Hard Look at Some Seductive Acronyms
    24 June 2004

    With the recent release of Intel's latest chipset, the community has been buzzing about a technology ordained to be "the next big thing" when it comes to storage performance - command queuing. How does the ability to retrieve requests out of the order they're received affect performance on the desktop, in the server, and across various RAID arrays?

    StorageReview has put several WD Raptor RAID arrays to the test against a comparable SCSI solution. You've got to check out the results- they will surprise you! TCQ, RAID, SCSI, and SATA
    quote:
    Hard Disk Command Queuing Goes Native in Seagate Drive
    By Mark Hachman
    September 16, 2003

    SAN JOSE, Calif.?Storage developers may get a look at a new hard disk technology here this week at the Intel Developer Forum. At the conference and expo, hard drive manufacturer Seagate Technology Ltd. will reportedly demonstrate a forthcoming mechanism incorporating "native command queuing."

    Seagate, based in Scotts Valley, Calif., is expected to announce the Barracuda 7200.7, a two-platter, 7,200-RPM, 200-Gbyte hard drive with 8 Mbytes of cache, sources close to the company said. Shipments of the drive are also expected to begin in November. Seagate will demonstrate a Serial ATA implementation of the drive at the Intel Developer Forum.

    Command queuing, as its name suggests, allows commands to be queued up, read and reordered before the driver acts upon data. The technology was defined as early as the ATA-4 specification, with the host controller determining the order of the commands to process. Native command queuing, part of the Serial ATA specification, allows up to 32 instructions to be queued and reordered by the hard disk controller itself.

    Meanwhile, Western Digital Corp. of Irvine, Calif., on Monday announced a new version of its Raptor 10,000-RPM disk drive, which uses the current version of command queuing left over from parallel ATA. The company will begin shipping this 10,000-RM Raptor drive this November, when the company will announce pricing.

    Western Digital's 74-Gbyte drive boasts the Ultra/150 CQ technology in addition to Rotary Accelerometer Feed Forward (RAFF) technology, designed to compensate for the vibration hundreds of drives produce while spinning in a drive cage.

    Seagate's implementation of native command queuing marks the beginning of a transition to a new technology. However, there are many performance considerations when comparing drives above and beyond their command ordering, such as the rotation speed of the platters, processor speed and the amount of cache.

    "They're both fairly similar in terms of what the drive does," said Mark Hartney, director of technical marketing with Silicon Image Inc., Sunnyvale, Calif., which provided components to both Western Digital and Seagate.

    Native command queuing requires a drive with fully integrated electronics, which can be "a little more efficient" than the type of queuing WD implemented, Hartney said. Western Digital's queuing, which the company calls "WD Ultra/150 Command Queuing," uses a bridge chip, he said.

    "The key difference between the two is the number of steps and the sequence of steps required for the queuing operation," said Knut Grimsrud, principal engineer on Intel Corp.'s I/O Architecture and Performance team in the company's R&D department; and chair of the storage industry's Serial ATA Working Group. The minimal command overhead implicit in the native Serial ATA implementation was the reason Intel chose to endorse it, he said.

    eWeek: Command Queueing

    SCSI Controller + IDE hdd? Tom's Hardware takes a look.

  17. #17
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb Raptor and ATA TCQ

    If you remember, the PCB of our WD740GD still has a Marvell 88i8030 chip, which means that WD740GD is an ATA drive, which can be connected to the SATA controllers because it is equipped with an ATA-to-SATA bridge.

    ATA protocol acquired support of requests queue processing a while ago, when they introduced ATA/ATAPI-4 specifications. But there was only one HDD manufacturer who implemented Command Queuing in the drives: IBM (see our article called The Last IBM Drive: Deskstar 180GXP HDD Details for details). This way Command Queuing in an ATA drive is possible not only theoretically, but also practically, and the hard disk drive manufacturers can definitely put it into life if they desire to.

    So, assume that the new Raptor supports ATA Queuing. However, in this case a question rises: does Marvell chip understand a set of commands implementing ATA Queuing?

    Let's check Marvell's site. This is what we find in the specifications for the chip (click here to go to Marvell's page):

    Supports ATA command queuing

    Well, this is exactly what I was striving at. But wait, and who said that CQ should only be supported by the drive? If the controller doesn't support queue processing then it doesn't matter at all if the HDD supports it or not. Tagged commands will never come to the drive in this case.

    Assume that the SATA controller we are using supports command queuing (since SATA 1.0/1.0a specification includes this support). Also we have a PATA drive on the other end of the cable connected via a special converter. In this case the SATA controller should work as a compatible device emulating the registers of a PATA controller. As a result, only initial PATA queuing can work in this case.

    [i]TCQ Part II: WD Raptor suprises X-bit Labs

    Before we start talking about the advantages of the new drives let's recall what we were unhappy with by the old 36GB Raptor:

    * No Command Tag Queuing support
    * Marvell serializing converter
    * Small storage capacity

    Compared with the first-generation Raptor drive, the new fellow features firmer and more massive stiffening ribs. But unfortunately, we again see the notorious Marvell controller on the electronics PCB.

    Western Digital announced that second generation Raptor drives acquired Ultra/150 Command Queuing (Ultra/150 CQ) technology. Is the so long-awaited revolution? In fact, the situation is not as simple as it might seem at first glance.

    I have every right to state that WD740GD drive doesn't process the commands with NCQ, i.e. it is not the native processing of the requests queue, which was officially announced in Serial ATA II: Extensions to Serial ATA 1.0a, revision 1.1. WD will switch from TCQ to NCQ in 2006.
    PDF manual

  18. #18
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb

    SATA II combines with SAS

    quote:
    External SerialATA Interface:
    First Facts about the Unannounced Standard
    by Nikita Nikolaichev

    Today we are going to talk about the external SerialATA standard, its peculiarities and differences from the internal SerialATA, as well as the problems the developers may face and the solutions they offer.
    ...

    As an example of the problems you might encounter when you connect external devices via the internal SerialATA version I could mention at least the fact that the standard warns against moving and shaking the interface cable during work. The acceptable signal level for internal devices and the cable screening requirements cannot ensure the desired protection against noises and distortions when you work with an external device. And the relatively weak connector used for internal solutions is beyond any criticism at all, if you decide to use it for external storage drives.

    All these things push you to the conclusion that external SerialATA requires different signal levels and a longer and better screened interface cable with a completely different secure connector.
    <snip>

    The requirements listed in the ?Serial ATA II Cables and Connectors Volume 2? specifications say that the cable for external SATA devices shouldn?t be longer than 2 meters and the cable itself should contain an additional integrated signal channel responsible for coordination of up to four external devices. This serial connection scheme is expected to be the most demanded for industrial and server devices, such as RAID subsystems, for instance.
    ...

    The connectors for external SerialATA devices have also undergone serious changes. In order to improve the protection against EMI (ElectroMagnetic Interference) and ESD (ElectroStatic Discharge) the developers paid special attention to proper screening. The requirements to mechanical specifications of the connectors have also grown tougher. The diversity of cables and connectors of the new type is determined by the above mentioned support of the multi-channel interfaces.

    A part of the specification related to the multi-channel connection of external devices, hasn?t been disclosed yet. However, the new standard requirements will be based on the SFF-8470 (Small Form Factor 8470) specification, which has been developed and applied for InfiniBand and Fibre Channel solutions. By the way, similar solutions were developed and will be implemented for more serious and demanding Serial Attached SCSI standard. In this respect, I would call the external SerialATA devices support a preliminary set of questions set and solved for the further use for Serial Attached SCSI standard.
    READ MORE: SATA Goes External

    quote:

    Group releases improved SATA standards -

    Improved Serial ATA standards are on the way.

    The Serial ATA II Working Group, which has been tasked with improving the Serial ATA standard, expects to release its findings by year's end. The new mark should see transfer rates of three gigabits per second. Among the new specs released were those for port selectors and port multiplier upgrades. The port selector spec allows two different hosts to connect to one drive, creating redundant paths. The port multiplier upgrade adds the ability for the multiplier to inform the host system if a drive is connected or disconnected from a port. The new specs are expected to create improved airflow and reduce cabling, as well as the obvious speed improvements.

    Though this is a great sounding technology it may get passed over. Home PCs may adopt this technology, but it will be very difficult to pull the server market away from SCSI. With the longevity of the SCSI standard and the future hopes of Serial Attached SCSI, this may be a short lived development. - by Brian Conant
    (September 22, 2003 - 11:56am EST) - www.geek.com
    Wiredstart
    quote:
    Serial-attached storage standardised
    SAS and S-ATA both standardised

    By Chris Mellor, Techworld

    Serial-attached SCSI (SAS) has become an ANSI standard. The http://www.scsita.org SCSI Trade Association (STA) and www.INCITS.org INCITS, the International Committee for Information Technology Standards, operating with ANSI approval, has approved SAS. The parallel SCSI bus technology is reaching the end of its life and a serial version of SCSI is poised to replace it.

    The specifications for the 3.0 Gbit/s SAS standard were in near-final form six months ago at which point OEMs began preparing products for market. By the end of 2003, several vendors had demonstrated working SAS products and are now beginning to announce product availability.

    The SAS standard is designated as ANSI/INCITS 376-2003 and can be purchased through the INCITS website. John Lohmeyer, chair of INCITS' Technical Committee T10 and a principal engineer for LSI Logic, said, "SAS preserves the industry's software investment in SCSI command sets. Its configuration is accomplished in software using globally-unique device identifiers. The new standard will make SAS products easier to design, install and maintain due to compatibility with earlier SCSI generations."

    The Serial ATA standard was approved in over a year ago, in November 2002, by the Institute of Electrical and Electronics Engineers (IEEE).

    In a neat two-step, parallel ATA and parallel SCSI are now both set to be followed by Serial ATA and Serial SCSI. Serial interconnects enable higher speeds and more devices than the shared bus ATA and SCSI designs. Drive manufacturers such as Western Digital, are already delivering SATA disk arrays. SAS arrays will surely join them throughout this year.
    Techworld

  19. #19
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb

    TCQ Part IV: mode page settings

    Now let's check if the diagnostics software will recognize CQ support by our Raptor drive. We will start with a very new and not very widely spread FC-IOMark utility (if you can think of a better name for it, your suggestions will be most welcome . We haven't yet finished working on it, but it already can disclose a lot of interesting things about the hard disk drives we run it on. We are particularly interested in the following section:

    WD740GD supports CQ with up to 32 requests queue depth! Well, here we can stop looking for the implemented command queuing support, as we have just found it

    Now let's try to figure out if the implementation of the CQ affected the WD740GD performance in any way. The easiest and the most illustrative way to estimate the CQ efficiency is to compare the performance of the drive with the enabled CQ support with the performance of the same drive with the disabled CQ support. As a result of this simple comparison we could get exactly this particular efficiency value, but we discovered some unexpected problems on the way.

    It turned out much harder to disable CQ support by WD740GD than to find it!

    On reading these words many of you could have recalled that most SATA controllers use SCSI mini-port as a driver. And it means that there is a SCSI Properties page, which appears in the HDD properties window. And on this page you can see Disable tagged queuing option. This all absolutely true: there is a page and there is an option.

    Of course, we selected this option immediately. However, our experiments showed that the HDD performance doesn't depend on the fact whether the Disable tagged queuing option is selected or not.

    This is unbelievable! But the surprises do not end up here. It turned out that selecting this option doesn't have any effect even for the SCSI HDDs! We tried to check how the thing works with an Adaptec 39320D controller and different hard disk drives, and suffered a complete fiasco.

    Well, it looks as if we will not manage to disable CQ in the SCSI Properties page. But let's not give in so soon. I have already had a chance to play a little bit with the supported queue depth on SCSI drives by editing the registry key (see our article called Ultra320 SCSI Interface: Highs and Lows. Part II for details):
    HKEY_LOCAL_MACHINE\System\CurrentControlSet\Servic es\adpu320\Parameters\Device\

    But then I simply changed the queue depth from 32 to 64 or 256. And if we set it to 1 or 2? Will this automatically disable CQ?

    Let's check this out now! We find a section of the registry where the SATA controller driver left a track (for instance if we use a Promise S150 TX2 Plus controller):

    Well, we see the magic word Tag and the number 33, which stands for 32+1. So, we change this parameter, run the benchmarks anew, and. Nothing changes! The HDD performance remained the same! Unfortunately, this method didn't work for any of the SATA controllers we had at our disposal. This parameter of the SATA drives seems to be used for some other purposes, I assume.

    Maybe we could try to estimate the CQ efficiency by comparing the dependence of the HDD performance on the queue depth? For this purpose we create a simple pattern in Intel IOMeter, which contains requests for random sectors reading, and run the benchmark a few times for different queue depths varying from 1 to 256 requests with 1 request step. We run the tests for two WD drives: WD740GD and WD360GD. If the dependence graph for the HDD performance under this type of workload will be shaped differently for both drives, then we can suppose that it is the effect of the CQ. So, what do we see?

    But even when we take a closer look at this given part we do not notice any principal differences in the dependence of the HDD performance on the workload. We see a flat area typical of the IDE drives under low workload, and a slight increase in the gap between the two testing participants as the workload grows. If this small gap is the only effect made by the CQ, then I should admit that I am pretty disappointed. But the results shown by WD740GD are most likely to be growing faster because of the faster actuator used in this HDD.

  20. #20
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb Spread Spectrum Clocking (SSC)

    Spread Spectrum Clocking (SSC)

    The chipset used by the SeriTek/1V4 and /1VE4 includes an Intel ASIC that doesn't support Spread Spectrum Clocking (SSC). This is a feature designed for SATA III. (We're still at SATA I.) There is even a debate as to whether SSC is even needed for SATA III.

    For some reason, Hitachi decided to be different from everybody and enable SSC in all their 7K250 drive. Hitachi decided to release an alternate version (and part number) for the 7K250s with SSC disabled. That version of Hitachi drive WILL work with the SeriTek 1V4 and 1VE4.

    Here is the list of part numbers that have SSC disabled and therefore do work with the SeriTek/1V4 and 1VE4:

    7K400 (400GB) - 0A30984 and 0A30985
    7K250 (250GB) - 0A30340
    7K250 (160GB) - 0A30339
    7K250 (120GB) - 0A30338
    7K250 (80GB) - 0A30337

    Current SATA 7K250 models:
    HDS722540VLSA80
    HDS722580VLSA80
    HDS722512VLSA80
    HDS722516VLSA80
    HDS722525VLSA80
    Photo
    All new Hitachi 7K250's have SSC disabled. T7K250, 7K400 and 7K500 White Paper

    Specification - OEM
    This document details the specification information for the Deskstar 7K250 (Serial) hard disk drive, model numbers: HDS722580VLSA80, HDS722512VLSA80, HDS722516VLSA80 and HDS722525VLSA80. Includes information on: Functional specifications (drive and disk characteristics, data integrity, electrical interface, signal timings, jumper settings, environment, power requirements, reliability, mechanical specifications, vibration, shock, acoustics, identification labels, safety, and electromagnetic compatibility) and interface specifications (registers, general operation, command protocol, command descriptions, and timings).
    Deskstar 7K250 Specification v1.5
    __________________
    Seagate 7200.9 family of drives ship with SSC enabled! and won't work until they release a non-SSC version, or offer a means to disable SSC.

    I assume that either this can't be fixed via firmware on controllers or from Seagate, and there is no SSC jumper block.
    Spread Spectrum Clock Generation (SSCG) is an efficient approach to reducing electro-magnetic interference emissions in many high-speed digital systems including hard disk drives.

    The most widely used conventional techniques to control electro-magnetic interference (EMI) emissions are shielding and filtering by using passive components. These techniques become difficult to employ when electronic systems become faster, more complex and portable.

    Shielding is the least desirable method of EMI reduction in most systems today, especially for portable and handheld products. That's because it increases the size, weight and cost of the system. Additionally, the increase in labor costs could be substantial, since shielding these products is difficult to automate in the manufacturing process.

    As for filtering, this technique is not systemic. Since designers cannot measure EMI emissions until after the system is built, they are forced to provide filter placement in many suspect locations and waste valuable time and PCB space.

    By contrast, SSCG offers good system-wide coverage as well as programmability, it does not degrade timing signal quality and it can reduce the number of pc-board layers and overall board space needed for a design.

    SSCG reduces the radiated emissions of the digital clock and timing signals by frequency modulating the system clock with a low frequency signal. This creates a frequency spectrum with sideband harmonics. Since the narrowband repetitive system clock is intentionally broadbanded, the peak spectral energy contained in the fundamental and harmonic frequencies is simultaneously reduced.

    All clock and timing signals derived from the spread spectrum clock are modulated at the same percentage amount. This leads to a dramatic EMI reduction throughout the system.

    . . .

    The difficulty in meeting the required emission levels is further increased when combining multiple hard drives in a single chassis, such as in work stations, since emissions produced by each drive is additive. EMI radiation can be easily controlled by programming the spread percent value up to +/-5.0 percent, depending of the system emission levels.

    In the case of hard disk drives used in desktop and notebook PCs, the spread spectrum technique reduce the number of board layers from six or four to two. In this case, the additional EMI reduction obtained by the use of Spread Spectrum Clock (SSC) eliminates the need for additional power supply and ground layers, resulting in only two-layer board, thus further reducing the system cost.
    Spread Spectrum Clocking
    Seagate pdf manual SATA 1.0 includes information on Spread Spectrum Clocking.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •