Results 1 to 7 of 7

Thread: eSATA 4+4 wierdness

  1. #1
    Join Date
    Sep 2005
    Posts
    44

    Default eSATA 4+4 wierdness

    Ok, just received two additional Seagate 300Gb 7200.8 drives from zipzoomfly and mounted them into spare hotswap cases for my Burly 4 box. Formatted and zeroed the 1st drive and monitored using IOSTAT and saw that it was averaging 47mb/sec during the 7-pass zeroing.

    After it was done (approx 12 hrs) I dragged it to the trash and attempted to format/zero the second drive (in the same Burly slot) and this time only got about 13mb/sec. 24hrs later it was still not done, so I cancelled it, and ejected it and put the 1st drive back in. Once again I attempted to zero/monitor this drive and lo and behold, it too was porking along at 12-13mb/sec.

    After rebooting the Mac the 2nd drive now zeros with approx 47mb/sec. Which leads me to believe that the Sonnet Controller "got confused". Needless to say, I'm not happy as there was no "error condition" -- just wicked slowness.

    Anyone else have similar experienc w/ eSATA cards?
    Any other insight as to what I was experiencing?

    Michael

    OS X 10.4.2, dual 1GHz MDD 2Gb RAM

  2. #2
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb

    This is I assume a continuation on your other thread about problems with Sonnet - I suggest you consider whether you stick by Sonnet or switch to SeriTek.

  3. #3
    Join Date
    Aug 2001
    Location
    Grangeville, ID USA
    Posts
    9,124

    Default

    TZ,

    Comon now, that is gettin silly. Changing from the Sonnet that is. The Sonnet is more capable than the Seritek and more compatible with new drives. Need to figure out what is happening but tossing the baby out with the bath water is not a solution. We are only seeing a 1 percent problem rate with the Sonnet card, been a very capable and worthy device. We are seeing much greater rates of weirdesses from the Seritek cards. I would have to strongly disagree with your recommendation here.

    Michael,

    Try a simple PRAM reset if you see something like this again. I am not quite sure why yet but several problems lately with SATA host cards and Disk Utility doing a zero all. A couple of customers have found a PRAM reset helped. It may be a problem in Tiger Disk Utility more than anything else because I can duplicate everything from Kernel Panics to drive dismounts with a multitude of drives and just about any SATA host card in several G5s and G4s. I am going to try and duplicate it with SCSI next and see if I can determine what is broken. Just about worn Brian out in the dungeon doing all these tests!

    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

  4. #4
    Join Date
    Jan 2001
    Location
    Mobius Strip
    Posts
    13,045

    Lightbulb

    I am basing what I said on AMUG review where they tested to see how well hot-swapping worked with SeriTek.

    And when it comes to synchronizing drives:
    Quote Originally Posted by ataMan
    ]You need an adequate controller which CAN handle the RAID-1 or RAID 1+0. A "normal" controller won't do the trick simply because there is no synchronization between user's actions and driver. Therefore during the hot-swap procedure it is impossible to distinguish between a drive in process of spinning up and a missing or failed drive. It's a non-issue on FireWire and SCSI because they involve a chain of drives on a single bus. On SATA we have several independent busses and the action on each bus needs to be synchronized. The SeriTek external controllers do the synchronization, but so far, no one else. It's unlikely it will change because FirmTek applied for patent. SATA Controllers
    And there is a commment further down...

    The Sonnet 4+4 is OK as long as you do not use hot-swap RAID (it works, but...), do not want to boot, do not have Sawtooth/Mystic and do not have Maxtor Maxline drives.

    Basically, if Sonnet is giving you grief three-times over, I would try another controller (I have boxes of most SCSI controllers from my search for the best/most reliable) in my own arsenal.

  5. #5
    Join Date
    Sep 2005
    Posts
    44

    Default dragging to trash first...

    whenever I hotswap out the drives, I always drag them to the trash first. Shouldn't that be enough to "force" the Sonnet eSATA card to release things?

    Also: this is *NOT* a RAID.

    Michael

  6. #6
    Join Date
    Feb 2003
    Location
    ex-Cupertino, CA, USA. Now HU+SRB (Europe).
    Posts
    163

    Default

    Quote Originally Posted by ricks
    TZ,

    Comon now, that is gettin silly. Changing from the Sonnet that is. The Sonnet is more capable than the Seritek and more compatible with new drives. Need to figure out what is happening but tossing the baby out with the bath water is not a solution. We are only seeing a 1 percent problem rate with the Sonnet card, been a very capable and worthy device. We are seeing much greater rates of weirdesses from the Seritek cards. I would have to strongly disagree with your recommendation here.
    Rick
    1) Personally you did say just the opposite many times and there are likely some of your posts on this forum of similar nature. Please stay behind of your own words.

    2) Please listen out these "new drives" (except: 500 Seagate issue which they already did fix - see my comment about it later). On the other hand, new Maxtor drives are still an on-going headache for Tempo cards - correct? Also, Sonnet cards are still not bootable and they (wisely) did stay away from intelligent hot-swap.

    3) Please report all the problems or "weirdnesses" to manufacturer. Otherwise if you don't - they won't know. As I know, I need only a single hand to count the number of 4-port cards you reported as "weird". More details about "weirdness" would be appreciated.

    No one says, Tempo family is not capable. In some cases it's purchase can be justified. In other cases it's too expensive, gives users what they do not need (extra 4 ports) but does not provide boot ability or intelligent, safe hot-swap. If a customer has a suitcase full of old Hitachi SATA-I drives and does not need hot-swap, staggered spin-up or boot from these drives - than 8-channel Tempo is the answer. Otherwise a question mark.

    Regarding Seagate: the 500 GB drives need to get SSC disabled regardless what controller they are connected to. With enabled SSC these drives do not work on G5 on-board SATA controller. Every drive purchased has to have the ability to be used also with the motherboard's controller for the safety. Unlike early Hitachi SATA-I drives, the SSC feature on Seagate 500 GB drives can be switched by software - and it's very fortunate.

  7. #7
    Join Date
    Mar 2003
    Posts
    203

    Default

    ataMan,

    I think I am going to have to be contrary to just about everyone's opinion in this matter.

    I have come to doubt that third party SATA implementations are sufficiently mature at this stage to recommend any of them to anyone who is not willing to throw a lot of hardware at a problem until some combination that works is discovered.

    I have been unsuccessful in resolving problems with a FirmTek 2+2 controller and it is beginning to appear unlikely that they will be resolved. As FirmTek are the only bootable SATA controllers available for the Mac at this time that alone is a very substantial limitation upon the usefulness of a third party SATA implementation. Then there are the problems of which drives various party's controllers will play well with and which drives are compatible with Mac implementations.

    Sure, the people doing HD Video capture don't have much choice. It's eight drive striped arrays or bust. The HD capture seems to be where Sonnet and some of the others believe their market is at the present time and hence they have not devoted resources to implement boot capability with their controllers.

    Are you sure this isn't "Nightmare on Elm Street"?

    It is just possible that PCs have a more mature implementation of SATA right now.

Posting Permissions

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