Oprava mrtveho HDD WD2500AAJS

[RayeR]

OK, no ja jen ze tam jsou jeste nake zavislosti na nakem jejich package manageru a bylo zminovano, ze treba ani v novejsi fedore to nemusi fungovat. Nasel sem na to tenhle navod:
https://computerlounge.it/how-to-unbrick-sandforce-ssd/
a je to snad nejdelsi how-to co jsem kdy cetl
Tak to gratulace, ze si na to mel trpelivost.
Jinak ten SF22xx v te dobe byl velmi popularni u ruznych vyrobcu SSD, dokonce i u low-end intelu, zrejme ze to v benchmarkach vypadalo dobre, ze diky te kompresi dosahoval zdanlive vyssich rychlosti...

[rnbw]

SandForce Package Manager je len generator image podla konfiguracie.

Moje poznamky:

Kód: Vybrat vše

Prepare firmware image(s): (should be run on Fedora Core 14 but Debian 9 works fine)
1. look up configuration number(s) in FW_Worksheet_SFI_SF-2000_140120_Confidential_Gold.xlsx according to controller and flash chips    -> e.g. 21795
2. edit lic file to include desired configuration number(s)
3. Use SF_OEM_PackageManager (from Genesis 1.7) to create firmware images & configuration:
 (before running, replace SF_Genesis with cracked one (from sfg.zip) to accept modified lic files)
        Distribution Firmware Package File:     FW_PPRO_5_8_2_Release.906190.release.dfp
        License File:                           edited lic file
        Customer Identification Number          0 (????)
        Diagnostics Unlock Key                  1,2,3,4,5,6,7,8,9,0,1,2,3,4,5,6 (????)
        Firmware Revision Override String       5.8.2
4. Use SF_ConfigurationManager to customize configuration:
        Model No.:                              OCZ-NOCTI
        Serial Number:                          OCZ-0000000000000000
        Fault LED Polarity:                     Inverted Polarity (or the red LED will stay on)
        Power Management Feature Selection:     B01

Flash firmware:
1. Connect SSD to PC with AHCI controller
2. Enter "factory mode" by shorting P4+P5 testpoints while powering up, then release
3. Use SF_Genesis to flash the firmware (as root):
        get address using lsscsi (e.g. 1:0:0:0)
        run SF_Genesis -PROCESS_SSD [DEVICE] [MFP_PACKAGE] [DRIVE_CONFIG_FILE] [SERIAL_NO] [WWN] [LOG=FILENAME]
        (e.g. ./SF_Genesis -PROCESS_SSD 1:0:0:0 FW_PPRO_5_8_2_Release.906190.release.0.mfp Configuration_Files/21795.0.driveconfig OCZ-0000000000000000 0000000000000000 log=log.txt)

Ale tvoj pripad to nemusi vyriesit. Ja som to mal uplne mrtve.

[RayeR]

Jo, dik za zestrucneny navod, muze se este nekdy hodit. To je uzitecne pro pripad, ze uz se SSD sam prepl do recovery mode a nejde s nim delat nic jineho, ja FW poskozeny nemam a slo to aktualizovat std cestou pres ADATA Field Updater. Ten take vypisuje nejake to signatue ID, podle ktereho se urcuje *.vic soubor.

Ted sem ale narazil na mnohem banalnejsi pricinu tech obtizi - teplota. Ted sem pred chvili zapl PC, kdy bylo vse studene a SSD normalne cetlo. Projel sem parkrat read test a benchmark a v poradku (je to na IDE kanalu, tak proto zadne stovky MB/s). Jakmile ale teplota dosahla asi 57C, tak doslo k drastickemu zpomaleni, ale este ne uplnemu zaseknuti. Sem to totiz docasne soupnul do IDE supliku, ktery nevetra, protoze mam odklopenou bocnici kejsu a tak uvnitr teplota rychle roste, kdyz skrz netahne vzduch. Na druhou stranu bych cekal, ze zdravy SSD vydrzi bez degradace vykonu nakych 85C nebo tak neco. Ono teda tu teplotu asi meri ten Sanforce cip a tezko rict, kolik maji flash chipy...
EDIT: OK, vyrobce pise 0-70C operating temp, no ale stejne, mel by tam mit nakou rezervu, asi ty pameti uz sou precejen nakople...

Zde par obrazku: