jamesbzx8

Member
  • Content Count

    61
  • Joined

  • Last visited

Everything posted by jamesbzx8

  1. Firewire port problems with Ultra ULT31310 external enclosure -------------------------------------------------------------------------------- There is this great deal at TigerDirect on the Ultra Enclosure. But the enclosures' Firewire port doesn't work flawlessly with every hard-drive: I have a Western Digital drive and a Seagate drive. Ultra advizes you to set the drives to MASTER with the appropriate jumper on the drives. The Western Digital drive wasn't recognized by Windows when using Firewire port. It was recognized when using USB port. There was no problem with the Seagate drive set to Master, as it was recognized by Windows when using either port. I then proceeded to swap the enclosure of the Seagate to put the Western Digital drive in it, this to verify what part was faulty. I put the Seagate drive in the enclosure previously used with the Western Digital. Surprise: the Seagate drive was recognized by Windows through the Firewire port, in the enclosure that the Western Digital drive wasn't beeing recognized. ( The Western Digital drive now in the enclosure previously used by the Seagate drive, was still NOT recognized through the Firewire interface in Windows) I know the WD drive is good because I can see it using USB and I have initialized it and formatted it through Disk Management in Windows. So I began doubting the jumper setting to set the drive to MASTER or SLAVE. On the PCB of the WD drive it is silkscreened where to put the jumper to set the drive as MASTER and this is what I followed. BUT On the sticker on the drive it shows an ALTERNATE way to set the drive as MASTER and for this alternate way, you don't need to install a jumper to set the drive as MASTER. I thus, removed the jumper and the Western Digital drive was then recognized through the Firewire port in Windows. Some years ago you also needed to look up what firmware your enclosures' bridge interface was at, but my problem wasn't of that sort. The bridge interface chip is the Prolific PL-3057 type In hoping this post will help plenty as I have scoured the internet for my problem and couldn't find a Googled result addressing my condition. Enjoy!
  2. jamesbzx8

    PSU Recommendations, please!

    Or just buy a cheapo PSU and use it to power the drives. Put the cheapo PSU lying at bottom of case if space is available. Leave PSU always on or connect a wire to the startup wire of your Antec with clip on connectors, no soldering of wires. Just an idea
  3. jamesbzx8

    RAID 5 File Copy Time

    Could you also specify wich controller you are talking about? Also give the configuration of the RAID arrays: stripe size, cache policies What is your platform hardware wise, memory speed on motherboard.
  4. jamesbzx8

    Question about RAID 5

    Run a regurlar backup to be put on an extra drive. of course this means you would only be backing just the essentials you can't afford to lose. You could get much lower performing, thus lower cost backup drives for this purpose. backing up the OS is of little utility because you can do a repair installation and have your system up in not much time. RAID5 or better 5E, 5EE is quite fast, little penalty on a hardware based RAID adapter. Cheap insurance is to make sure your drives have dedicated cooling fans pushing air over them. Some ideas here to help you think things over JB
  5. jamesbzx8

    RAID Guru help please....

    Hello, bonjour ;-) I have a ServeRAID-6M at home on a limited 64bit/66MHz PCI slot. This is on a home machine, and I have tried various configs with 6 disks, all 15kRPM Fujitsus MASxxxx. Is your RAID5 a 5E or 5EE? What is your stripesize,Readcache and writecache settings? Are there any background task running on your array/controller? For your environment (SQL) you should bench/ characterize your setup with IOmeter or Intel IPEAK Storage Performance Toolkit (not supported by Intel now) tweakers.net use this kit and have many load profiles that would help you get a better grasp of what works for your setup. You could also try to find hints/help at: http://www.sql-server-performance.com/ http://www.enterpriseitplanet.com/ http://www.enterprisestorageforum.com/ you could look at IBM, on SQL tuning tips: http://www.ibm.com/Search/?q=sql+tuning&am...c=us&en=utf Check this old discussion on ServeRAID-6M performanbce: http://forums.2cpu.com/showthread.php?t=57...ht=serveraid-6M I hope this will help you find a solution or give you ideas on finding your solution, JB
  6. I have just acquired this RAID controller, an IBM ServeRAID-6M with 256MB cache. It had 6.0x BIOS and firmware level. I proceeded to plug the controller on a iWill DH800 motherboard with no cables and drives connected. I wanted to see it POST, and it did and all appeared normal. I downloaded the "application CD" image and the "Support CD" image from IBM support pages, they are version 7.12 kit. I updated BIOS and firmware level with this and everything went well. I went into the mini-config utility of the card with CTRL-I and saved controller default settings and rebooted the controller. All well again. I then proceeded to install a 4 device SCSI-320 braided and terminated cable to the controller channel-1 and connected 2Fujitsu MAS3367. The controller wouldn't recognize any drive at all. It would spend 6 minutes at the drive configuration stage of POSTing and It would report that : 1:ServeRAID-6M slot 4, Logical drv=0, Firmware=7.12.11,Status=Fail 1 device(s) not responding Controller OK; drives not responding check, cables or enclosures. Press F2 Detailed information F4 Retry the command F10 Continue without changing the configuration I tried removing the battery from the controller--> same result. I tried re-saving BIOS default--> same result I tried another SCSI-320 cable--> same result I tried disabling spin-up mode on my drives---> same result I go into the IBM ServeRAID mini-configuration utility with CTRL-I and go to controller status screen and it reports: controller config is invalid. With registers: ISPR=EF10h BCS=01h ECS=00h I am lost, I have other RAID controllers and have never had any problems having drives beeing acknowledge by them. I have eXtremeRAID 2000, ServeRAID-4H, MegaRAID 1600 Enterprise, AcceleRAID 600 and LSI 320-2X and all work great. Can someone help, am I missing a jumper on the controller or does the controller need IBM disks or does the controller need special cables. Thank you for your time or referring someone esle to this forum to answer my plea JB
  7. jamesbzx8

    ServeRAID-6M can't recognize drives

    Hello again, Burned myself a "ServeRAID Support CD v8.30 " and a "ServeRAID Application CD v8.30 ". I booted with the v8.30 Support CD and since the controller BIOS/firmware doesn't detect any drives attached to any channels, it didn't provide me with any new option. I can add that the ROM Update Wizard in that version won't update a ServeRAID-6M also since they came out with this version for the 8i/8k/8k-l series controller support and this is shown in the IBM ServeRAID Software Matrix. I then thought that I would try the "IBM Hard Drive Update Program (Windows package) " and took out my Adaptec 2940u/UW to do this but I stopped and thought of a few more things to try with the ServeRAID-6M. There is a jumper with the mention FLASH beside it, I tried to look up any info of this in the documentation but to no avail. Still I proceeded to install a jumper on it and powered up the machine. The whole row of RED LEDs lit up and this made the boot-up process skip the POSTing of the card, like it disabled the BIOS of the card. I then powered down the machine and looked at the Timer block wich is the yellow cube underneath those just mentionned RED LEDs and beside the black connectorthat has silkscreened next to it JTAG. I looked at its underside with a jewellers monocle/magnifier and saw that I could remove that yellow top. This yellow top is a Lithium battery that powers the chip/IC under it ( I'm learning new stuff here, I had never seen that arrangement) I felt lucky as I was pretty sure I would erase totally the BIOS, except for the boot block, I then replaced the lithium battery after taking it off for about 30 seconds and I also removed the previously installed jumper. After re-seating the controller on the motherboard, I powered up the system and the ServeRAID-6M POSTed and reported that there was an irrecuperable configuration failure and if I wanted to continu on or press F9 to reset its configuration to a default state; I pressed F9, of course I restarted the machine and it was the sad same behaviour as if I hadn't done anything I was resigning to repack the card and make arrangements for its trip back to the seller. As I was removing the card from my machine, I said to my self that I should make a thourough Visual Inspection of it with my jewellers monocle/magnifier hoping to maybe find a strand of conducting fine wire or a speck of metal or solder, anything that could create a short, solid or intermittent. I hope you are still with me here and that this full account is entertaining and maybe even instructive. Drum roll please --> I found three damaged surfacemount components. They are near the faceplate/backplate of the card on the backside/solderside where there is less parts on. It appears that the seller has damaged, mishandled the card before puting it in the antistatic bags and in the shipping box before sending it out to me . I looked in the antistatic bags and there was no parts in them. If you have a ServeRAID-6M next to you please have a look at these three parts: L6 inductor ( black rectangular) R158 resistor (very small black rectangular) C67 capacitor ( very small beige rectangular) -The resistor was hanging by one side and I succeeded in replacing it and resoldering it. -The capacitor was very damaged and I evaluated that the card might work without a perfect one replaced. -The inductor was nowhere in whole or in part, but by the size of the two surrounding similar ones I figured a lenght of small gauge of stranded wire would do the trick in its place. After this surgery, I re-seated the card in my machine and it worked , it recognized my attached drives to both channels. I am trying it out with a RAID-0 array and a RAID-5 array. The controller appears to test right If it performs without any glitch I would consider keeping it. The seller fully recognized my efforts, didn't have another one to exchange it for and very gentlemanly agreed to compensate me for my work! Another eBay transaction going well JB
  8. jamesbzx8

    SCSI U160 controller

    Just want to correct that it is possible to run a IBM ServeRAID-4H on WinXP but you have to edit the .inf in the driver setup disk so that the drivers are taken by WinXP. By default still, WinXP installs an old version of drivers for the ServeRAID-4H
  9. jamesbzx8

    RAID 10 slower than RAID 0?

    You could get a U160 enterprise class caching RAID controller for not much on eBay. Some of them work in a 32bit PCI slot. The ones that work are eXtremeRAID 2000, ServeRAID-4H and MegaRAID Enterprise 1600. They are 4 channels U160 controllers, so you could put 2 drives per channel and not feel much bottleneck except for the burst to U320. Maybe it would be an inexpensive try. They also have all the high-end cabalities like hot-sparing, capacity expansion, some RAID migration and rebuild in background. Some ideas for you JB
  10. jamesbzx8

    IOMeter patterns

    do a search with IPEAK, IIRC it has a function to record the disk access paterns you assign it to. For IOmeter, go to tweakers.net, they use it regularly, but iti is in dutch methinks. You can do a search on theuir site or contact the editors of the site. I downloaded from them all of their test patterns for IPEAK/IOmeter. But I have deleted all of that stuff a while ago ;-(
  11. jamesbzx8

    RAID 10 slower than RAID 0?

    I would think that the MB/s traffic of data to the mirror drives would have some sort of penalty on the bus they're on, the cache allocation( caching controller) and some management of this traffic and SCSI commands by the RAID controllers' processor. Of course I agree that calculating parity is a task that occupies the processor ( wether software RAID or dedicated RAID processor)
  12. Very good observation Discjet! My assumption is that it is a hybrid backplane and probably expects to receive a stand-alone terminator on the back 68 pin connectors. If the initial poster would reveal the brand of this backplane/enclosure, it would cut down on assumptions ;-) Thanks for participating :-)
  13. Hello, I am presently toying with various Enterprise class caching RAID controllers, mostly U160 ones and a couple of U320 ones. I am also using 6 Fujitsus' MASxxxx and 3 MAUxxx. I also get access times like you are getting with HDTach, around 5.4-5.6mS. I would definitely change your cable wich is one of your source of frustration. I have toyed with an old semi-rigid Ultra cable, its' termination was choking my drives. I then employed proper semi-rigid BRAIDED , terminated U320 cables. These BRAIDED cables are from FoxConn and have connectors for 4 devices. Igot them from STSI.com from eBay. They are working without a hitch. They are reasonably priced around 15-18 dollars $US. I have no idea if your Hybrid SCSI backplane case is of sound design. You might also want to look at your motherboards' BIOS setting that could bottleneck your setup. This setting is called PCI Latency Timer: older motherboards used to have this set at 32. You should try to have it at 64: newer mobos have this value now set by default in the BIOS. Here is a linky to the Great Adrian Rojaks' BIOS setting guide: http://www.adriansrojakpot.com/Speed_Demon...S_Guide_02h.htm Another possibly overwhelmed device might be the 29160, its processor was designed long time ago and having no cache, might be further limiting STR if you RAID0 the drives you have. This HBA ( Host Bus Adapter) came out when SCSI drives weren't speeding demons like the nice 15K drives you have. I hope that my suggestions help you in your quest and also that you don't feel all alone ;-) Enjoy! JB
  14. jamesbzx8

    Java start and file download

    Like Eugene mentioned up above, I'm one of the savvy forum goers who knew better than to click that VIP staus toolbar link. And then yesterday, while visiting here, Kaspersky 6.0 saved my PCs *ss and mine ;-) by catching the malicious trojan scripts installment. Thank you Admins for taking care of this quickly and revising the security of this site!
  15. jamesbzx8

    Movin Array to New Controller

    standing by to read more of your findings ;-)
  16. jamesbzx8

    Setting up a server, SCSI Raid options...

    Maybe ServeRAID-4L is a 64bit/33MHz card. I have the 4H and it is speced at 64bit/33MHz and same as for you there is barely any measurable difference between 33bit and 64bit slot, mind you this is just with 2disk RAID-0
  17. I agree with the above, the MFT must be damaged: Master File Table (MFT) An NTFS system file on NTFS-formatted volumes that contains information about each file and folder on the volume. The MFT is the first file on an NTFS volume. I believe there are some utilities also taking care of this. I have used EasyRecovery PRO and its' work is excellent
  18. I too, am wondering if the board was not underclcoking itself because of heat issue after running overclocked Is the original poster still here?
  19. jamesbzx8

    Need your opinion people

    Seems last poster is right, Have a looky over here the almost same discussion is on there: http://forums.2cpu.com/showthread.php?t=77022
  20. jamesbzx8

    Need your opinion people

    Did you try with Writeback cache setting? Is there a setting on your motherboard that would limit the PCI-e slot that the controller is in? Have you looked in BIOS for the setting PCI Latency : 32 or 64, 64 beeing now prefered. Did you try with stripe size set at 64 on your RAID? These are some settings that vary my results to various degrees for me.
  21. jamesbzx8

    Slow SCSI-configuration

    Did you try the connecting the harddisk to the empty connector next to the terminator on the cable?
  22. jamesbzx8

    Raid 5 with sleep mode

    Please find information on power management at this adress: http://book.itzero.com/read/microsoft/0507...09lev1sec5.html Also my personnal experience is that I have in the past thought that my system was frozen because I lost patience waiting for the computer to wake-up,and was reseting my 'puter. But, it was just a long waking-up process and the system eventually got out of its sleep state. As the article shows you what to look for in device manager, experiment with your system, you have nothing to lose but time itself. Also, as the article spells out, there is a walk-through the devices that the PowerManager does and this process can give the appearance that the computer is frozen: arm yourself with patience as you wait during that time when summoning the computer from its sleeping state. On my Mylex eXtremeRAID 2000, it turns off the controller in all of the power saving state. Enjoy!
  23. jamesbzx8

    Heat transfer foam

    I believe this foam is called thermofill, look for it under that name
  24. That's what I used yesterday to remove lots of left-over adhesive from the bezel of an IBM 21" monitor I bought at a surplus store. Did it ever work well--> Good ol' WD-40 I first tried paint thinner Sweet Mother of Jesus, did it ever remove the adhesive and the plastic of the bezel also --> Ouch! Fortunately I was trying it on the side-bottom of the bezel So do not forget, WD-40 is the magic remover ( and it leaves a nice smell )
  25. jamesbzx8

    What's the fastest processor available?

    Stop complaining why people don't answer your question the way you like. Try to find the useful bits and appreciate what you got, especially when it was a OFF-TOPIC question to begin with and you failed in your own research. True, so true