Skip to content

Hardware RAID Controller Failure: Amannɔne Configs a wode bɛba a wuntumi mpopa wo Drives

Ɛsono dodow a wɔde ma ntɛm ara wɔ ɔman biara mu. Kratafa yi de 911, nɔma a wɔde ma United States ne Canada na edi dwuma. Aman afoforo pii de 112, 999, anaa nɔma foforo di dwuma. Hwehwɛ nɔma a ɛwɔ wo man mu.

Bere a adwumakuw bi somfo hardware RAID controller kaad, te sɛ Dell PowerEdge PERC (H730P, . H740P, . H755, . H840), HPE ProLiant Nyansahyɛ Nhyehyɛe (P408i-a, . P440ar, . E208i, . MR416i), anaa Broadcom / LSI / Avago MegaRAID (SAS 9361-8i, . 9460-16i, . 9560-16i), ohu amane wɔ PCIe board wu, BBU/CacheVault supercapacitor huammɔdi, anaasɛ mpofirim backplane voltage drop, na wo install RAID card a wɔde besi ananmu nkutoo na wuhu sɛ pre-boot BIOS utility pause ne ehu asɛm "Foreign configuration(s) found on adapter. Press any key to continue or 'C' to load the configuration utility", keystroke biako a ɛnteɛ betumi asɛe terabytes a ɛyɛ production databases ne virtual machines wɔ nea ennu sikani abien mu. Wɔ hardware RAID architectures mu no, array no stripe geometry (RAID level, drive nhyehyɛe, . 64 KB / 256 KB stripe size, ne Logical Volume LBA hye) wɔkora abien no nyinaa wɔ controller card no onboard NVRAM chip ne wɔ hidden reserved sector wɔ honam fam SAS/SATA/NVMe disk biara a wɔfrɛ no Disk Data Format (DDF / . COD, Nsiesiei wɔ Disk so) metadata mantam. Sɛ RAID kaad foforo koraa a wɔde besi ananmu (anaasɛ kaad a ɛwɔ hɔ dedaw wɔ bere tiaa mu backplane tumi glitch akyi) boot up na ehu drive ahorow a ne on-disk DDF metadata serial number no ne card no NVRAM a ɛda hɔ kwa no nhyia, ɛhyɛ saa drive ahorow no frankaa sɛ ** ."Foreign"** na ɛyɛ. Sɛ ɔhwɛfoɔ a ehu aka no klik "Create New Virtual Disk" anaa "Initialize" a ɔsusu sɛ "ɛma" drives no, anaasɛ ɔbɔ "Clear Foreign Configuration" bere a na wɔpɛ sɛ wɔde ba a, controller no kyerɛw on-disk metadata no so na ɔde zero fi Virtual Disk headers no so. Nsonsonoe pɔtee a ɛda ** ntam a wonim .Import Foreign Configuration** ne ** .Clear Foreign Configuration**, sɛnea wɔkyerɛw honam fam drive bay nɔma ansa na wode wo nsa aka caddy biako, ne sɛnea wodi Pinned / Dirty Write-Back Cache (BBU / CacheVault) ne "Punctured Stripe" multi-drive dropoffs bɛma wo array no asan aba intanɛt so a data ayera biara. Sɛ anyinam ahoɔden ogya a ɛtɔe wɔ server dan mu, UPS lithium-ion thermal runaway, anaa nsu paipu a ɛpaee na ɛde server no dii huammɔ a, fi dan no mu na frɛ wo mpɔtam hɔfo nɔma ahorow a egye ntɛmpɛ (911 wɔ United States ne Canada, 999 wɔ United Kingdom, 112 wɔ Europa Aman Nkabom no nyinaa mu) ntɛm ara. Sɛ wopɛ storage array, virtual machine, ne SSD ntɛmpɛ a ɛfa ho a, hwɛ yɛn akwankyerɛ ahorow a ɛfa ho NAS RAID array asɛe, draibɛ pii adi nkogu, ne ZFS/Btrfs sanba, . virtual machine VMDk, VHDX, ne QCOW2 snapshot nkɔnsɔnkɔnsɔn asɛe, na SSD NVMe mpofirim akenkan-nko mode ne spare blocks gye.

Yɛ eyi mprempren

  1. MMA KLIK "Initialize", "Fast Initialize", anaa "Create New Virtual Disk" da wɔ drive ahorow a data a ɛwɔ hɔ dedaw no so. Wɔ Dell PERC, Broadcom MegaRAID, ne HPE Smart Array nsɛmfua mu no, Initialize ENNYƐ "dan na fa disk no si so" (te sɛ nea ɛtɔ mmere bi a ɛkyerɛ wɔ legacy macOS/Windows disk utilities mu). Wɔ hardware RAID firmware mu no, ** .Initialize kyerɛ sɛ FORMAT AND ZERO OUT THE VIRTUAL DISK METADATA AND PARITY BLOCKS** na ɛyɛ. "Fast Initialization" a ɛyɛ sikɔne 3 mpo kyerɛw megabyte a edi kan ne nea etwa to wɔ logical volume no so, na ɛpopa GPT mpaapaemu pon, VMware VMFS asɛmti, Windows NTFS Master Fael Table boot sector, anaa Linux LVM2 / XFS superblock a ɛyɛ nwonwa.
  2. Yi agyiraehyɛde a ɛtra hɔ daa mprempren ara na kyerɛw Physical Slot / Bay Number (Bay 0, . Bay 1, . Bay 2...) wɔ drive caddy biara so ANSA na wobɛtwe disk biako. Bere a nnɛyi SNIA DDF metadata sie drive sequence tags wɔ disk biara so, firmware dedaw mu nsakrae, backplanes a wɔakyekyɛ mu, anaa nsaano ntɛmpɛ a wɔsan yɛ (No-Init Recovery) de wɔn ho to disk biara a ɛsan kɔ ne mfitiase pɛpɛɛpɛ so ɔha biara mu nkyem 100 Enclosure:Slot bay nɔma. Mma ntwe drive anan mfi server bi mu da na mfa nhyɛ pon so a wɔmfa nkyerɛwde biara nka ho.
  3. Ntwe apɔwmuden anaa "Rebuilding" draibɛ da na san fa hyɛ mu bere a server no wɔ ahoɔden sɛ "sɔ hwɛ sɛ ɛsan twitwiw kɔ soro." Ntɛm ara a wobɛtwe SAS/SATA caddy a ɛte ase afi hot-swap backplane mu bɛboro sikani 5 no, RAID controller no hyɛ saa disk no bere nsɔano agyirae sɛ Failed / Offline anaasɛ Foreign. Sɛ na drive biako awu dedaw wɔ RAID 5 array mu na wode anifere twe drive a ɛto so abien fi mu kɔhwɛ ne part number bere a server no reyɛ adwuma a, wokum drive a ɛto so abien no ara kwa wɔ controller no ani so na woabɔ RAID 5 volume no nyinaa hwehwe.
  4. Te mmara a ɛho hia a ɛfa Import vs. Clear: Sɛ woresesa RAID kaad a awu (anaasɛ bere a drive ahorow nyinaa abom atɔ offline bere tiaa bi akyi) a, wopɛ IMPORT. Sɛ wode spare drive a wɔde adi dwuma bɛhyɛ server a ɛretu mmirika a ɛyɛ papa mu a, wopɛ CLEAR.
  5. Scenario A (RAID card no wui na wode card a wɔde besi ananmu sii hɔ, ANAA power/backplane glitch bi maa wo array no nyinaa kɔɔ Foreign): Nokware nhyehyɛe no te ase wɔ disk ahorow no so (Configuration on Disk) na ɛyɛ. Ɛsɛ sɛ wopaw ** .Import Foreign Configuration** enti kaad a wɔde besi ananmu no kenkan array nhyehyɛe no fi disk ahorow no so kɔ ne NVRAM mu.
  6. Scenario B (Wo server no reyɛ yiye, 1 drive awu, na wo slid replacement drive a wɔde adi dwuma a wɔatwe afi server dedaw a wɔakɔ pɛnhyen mu akɔ Bay 4): Saa replacement drive a wɔde adi dwuma no wɔ RAID metadata dedaw a efi server a wɔakɔ pɛnhyen no so wɔ so (Foreign). Sɛ wobɔ so a Import, wobɛfa server dedaw a wɔakɔ pɛnhyen no ghost array no aba. Wɔ Scenario B mu no, wopaw ** .Clear Foreign Configuration** wɔ saa disk biako a wɔde besi ananmu no so sɛnea ɛbɛyɛ a ɛbɛyɛ Unconfigured Good na ofi ase san si sɛ ade a ɛyɛ hyew a wɔde si hɔ.
  7. Sɛ drive mmienu anaa nea ɛboro saa tow offline wɔ RAID 5 (anaa abiɛsa wɔ RAID 6) mu a, MMƆ drives no wɔ intanɛt so na fa "Rebuild" ansa na woayɛ cloning drive biara ddrescue. Sɛ drive ahorow pii fi array bi mu esiane sectors bɔne nti ( .UREs, Unrecoverable Read Errors), sɛ wohyɛ draibɛ a akyɛ sen biara a wɔatow agu no san kɔ intanɛt so a, ɛma "Stale Data Desynchronization" anaa "Punctured RAID Stripe" a ɛsɛe database ahorow daa. Fa power down server no, fa drives no bata HBA a ɛnyɛ RAID ho (IT-Mode SAS card), clone drive biara bit-for-bit kɔ mfonini fael ahorow mu, na ɛkame ayɛ sɛ san yɛ array no wɔ software a wɔkenkan nkutoo mu kan.

Danger

Nneɛma a "Pinned / Dirty Cache" (CacheVault / . BBU) Data Porɔw Afiri Bere a Woresesa RAID Kaad a Awu. Enterprise RAID controllers de 1 GB kosi 8 GB a ɛwɔ onboard DDR4/DDR3 Write-Back Cache RAM a supercapacitor gyina akyi di dwuma (CacheVault) anaa Lithium-Ion Batere a Wɔde Siesie Nneɛma (BBU). Sɛ wo server no hwe ase wɔ database a emu yɛ duru mfinimfini anaa VMware kyerɛw esiane sɛ RAID kaad no PCIe logic chip no wui nti a, ebia gigabytes pii a ɛnyɛ nea wɔahyɛ da ayɛ fael-system metadata a ɛte kaad a awu no CacheVault flash module anaa daughterboard RAM a wotumi yi fi hɔ no mu (Dirty / Pinned Cache) a antumi ankɔ hard drive ahorow no so da. Sɛ woresesa Dell PERC anaa LSI MegaRAID kaad a awu a na ɛwɔ Write-Back Cache a ɛyɛ adwuma a, hwɛ wo hardware service manual no mu hwɛ sɛ CacheVault flash module / NVRAM daughterboard no yɛ nea wotumi yi fi hɔ na wobetumi de akɔ kaad a wɔde besi ananmu no mu (anaasɛ wobɛhohoro cache a wɔakora so no ansa na wode aba) de asiw ano NTFS / . VMFS / . ext4 nsɛmma nhoma mu adifudepɛ.

Gyinaesi dua: Hardware RAID controller a wɔde besi ananmu ne Amannɔne Nsiesiei triage

flowchart TD
    A["Hardware RAID Emergency: Controller Card Awu ANAA BIOS Bɔ Amanneɛbɔ sɛ ‘Wɔahu Amannɔne Nsiesiei’."] --> B{"Dɛn nti na ‘Amannɔne Nsiesiei’ kɔkɔbɔ no reda adi?"}
    B -->|"Asɛm 1: Wɔasesa Dead RAID Card (anaasɛ wɔde drive nyinaa akɔ server chassis a wɔde besi ananmu) ."| C["Honam fam Drives Nyinaa Yɛ Akwahosan & Kura Wo Ankasa Array Metadata ('Nsiesiei wɔ Disk so')"]
    B -->|"Asɛm 2: Backplane / PSU Glitch maa Drives pii a ɛwɔ Server a ɛwɔ hɔ dedaw mu no hwee ase kɔɔ ‘Amannɔne’ bere koro mu"| D["Power Down Server -> Reseat SAS Backplane Cables & Drive Caddies (Wɔ Pɛpɛɛpɛ Numbered Slots) -> Boot to Controller BIOS"]
    B -->|"Case 3: Wode Replacement Disk 1 a wode adi dwuma no ahyɛ Running Array mu Nkyɛe & Saa Disk 1 no Ka sɛ 'Foreign'."| E["MMFA MMFA MMƆ: Saa disk 1 a wɔde adi dwuma no kura metadata dedaw a efi server foforo mu; paw 'CLEAR Foreign Config' sɛnea ɛbɛyɛ a ɛbɛyɛ Hot Spare"]
    B -->|"Asɛm 4: Drives pii a wɔrekliki honam fam / adi nkogu wɔ Media Mfomso (RAID 5 a ɛwɔ 2+ Bad Disks) ."| F["GYINA: MMA Nhyɛ Intanɛt So Anaasɛ San Si; Fa Drives bata IT-Mode HBA ho, Clone Disks Nyinaa denam ddrescue so, & Ɛkame ayɛ sɛ Reconstruct"]
    C --> G["Hyɛn Controller BIOS (iDRAC / Ctrl+R / F2 System Setup) -> Bue 'Amannɔne Nsiesiei Hwɛ / Nhwɛso'."]
    D --> G
    G --> H{"Hwɛ Amannɔne Preview Screen no: So Virtual Disk no member drive NYINAA wɔ hɔ na ‘Online’?"}
    H -->|"Yiw: Drives nyinaa (anaasɛ N-1 wɔ RAID 5 mu / N-2 wɔ RAID 6 mu) wɔ hɔ"| I["Yɛ 'IMPORT FOREIGN CONFIGURATION' (MMA KLIK Clear anaa Initialize DA); Virtual Disk no Ntɛm ara na Ɛforo a Ɛnyɛ hwee"]
    H -->|"Dabi: Drive pii da so ara kyerɛ ‘Missing’ wɔ Preview no mu"| J["ABORT IMPORT: Mfa stripe set a enni mũ mmra; shut down na diagnose honam fam drive/cable bɛn na entumi ntwa nkɔ soro"]

Adetɔnfo-ne-Ɔtɔnfo Nkyerɛkyerɛmu: Sɛnea Dell PERC, Broadcom/LSI MegaRAID, ne HPE Smart Array di Amannɔne Metadata ho dwuma

Adwumayɛbea server tɔnfo kɛse biara de nsɛmfua soronko kakra di dwuma wɔ ne UEFI HII nhyehyɛe, Out-of-Band Management (iDRAC / . iLO / . XClarity), ne CLI adwinnade (perccli / . storcli / . ssacli). Hwɛ adetɔnfo ntotoho pon yi pɛpɛɛpɛ ansa na woamia safoa biako so:

Feature / Adeyɛ Dell PowerEdge PERC (H730, . H740P, . H755, . H840, a egyina LSI so) . Nsɛm a Wɔka Kyerɛ / LSI / Avago MegaRAID (9361, . 9460, . 9560) . HPE ProLiant Smart Nkyekyɛm (P408i, . P440ar, . E208i, Microchip/Adaptec a egyina so) .
Baabi a Array Metadata Te wɔ Disk Biara So SNIA na ɔkyerɛwee DDF (Disk Data Format) a wɔde asie wɔ awieɛ 64 MB kɔsi 512 MB wɔ honam fam SAS/SATA/NVMe disk biara awieeɛ pɛɛ. SNIA na ɔkyerɛwee DDF (Disk Data Format) a wɔde asie wɔ final sectors a ɛwɔ honam fam disk biara awiei. RIS (Reserved Information Sector) metadata a wɔde asie wɔ honam fam draibɛ ahorow no so (abakɔsɛm mu wɔ draibɛ no awiei wɔ Smart Array so P-a ɛtoatoa so).
So Wubetumi Asesa Dead Card ne Model Foforo Kakra? YIW: PERC biara a egyina LSI so (H330, . H730, . H740P, . H755) betumi akenkan na wɔde aba DDF Amannɔne Nsiesiei a ɛfa PERC foforo a egyina LSI so (bere tenten a kaad foforo no boa RAID level ne drive count koro no ara). YES: Forward-compatible wɔ MegaRAID nyinaa so SAS 2108 / 2208 / 3108 / 3508 / 3916 Tri-Mode controllers nam gyinapɛn so DDF Amannɔne Nneɛma a Wɔde Ba amannɔne. YES: HPE Smart Array controllers wɔn ankasa kenkan on-disk no RIS metadata wɔ boot a enhia mpo sɛ wode nsaano "Amannɔne Import" klik wɔ nsɛm dodow no ara mu (anaasɛ bisa Press F1 to keep configuration on drives vs F2 to erase).
Ahyɛde / Menu Pɛpɛɛpɛ a Wɔde Gye Array no ** F2 System Setup -> Device Settings -> PERC -> Foreign Configuration Management -> Preview -> Import Foreign Configuration (anaasɛ ɛnam ** soiDRAC Storage -> Controllers -> Import Foreign Config**). ** Ctrl+R / UEFI HII -> Foreign View -> Import (anaasɛ ɛnam CLI so: storcli /c0/fall show preview akyi no, ** .storcli /c0/fall import**). Wɔ POST nsɛm a wɔka kyerɛ no so: 1789 - Slot X Drive Array Disk Drive(s) Respond... Press F1 to continue with logical drives enabled (MMIMIA DA F2 wɔ HPE POST asɛm bi so, sɛnea F2 gye data a ɛyera / popa config).
Fatal Button a Ɛpopa Biribiara Sɛ wobɔ ** a.Clear Foreign Configuration** (ɔpopa DDF metadata a efi disk ahorow so) ANAA ** .Initialize Virtual Disk** (zero fi logical drive no mu). Mmirikatu ** .storcli /c0/fall del** (popa amannɔne config) ANAA ** .storcli /c0/v0 start init** na ɛyɛ. Sɛ wobɔ ** a.F2** bere a POST bere a drive bi ayera bere tiaa bi ("1785 - Drive Array Not Configured") anaasɛ wobɛbɔ ** .Re-create Logical Drive ne Mfitiaseɛ**.

Playbook 1: Anamɔn-anamɔn ahobammɔ "Amannɔne Nsiesiei Import" wɔ Dell PERC & LSI MegaRAID

Bere a wo Dell PowerEdge anaa Supermicro/Lenovo MegaRAID server no gyina wɔ boot ne "Foreign configuration(s) found on adapter", di oprehyɛn ho nhwehwɛmu kratasin a ɛwɔ anammɔn 3 pɛpɛɛpɛ yi akyi:

Anamɔn 1: Hwɛ Physical Drive LEDs ne Spin-Up Gyinabea Di kan

Ansa na wobɛkɔ BIOS menu no mu no, nantew kɔ server chassis no anim (anaasɛ bue Dell iDRAC -> Storage -> Physical Disks) na kan drive bay LEDs no: - Adɛn nti na array no bɛyɛɛ Foreign wɔ nea edi kan no mu? - Sɛ woasesa PERC/MegaRAID kaad a awu kɛkɛ (anaasɛ wode draibɛ nyinaa akɔ spare server chassis mu), a wubehu Foreign wɔ all drives so no yɛ ɔha biara mu nkyem 100 normal na wɔhwɛ kwan. - Nanso, sɛ wo annyɛ sesa RAID card no na server no rebooted mpofirim mu "Foreign configuration found", ɛkyerɛ sɛ wɔtow draibɛ pii agu offline bere koro mu, mpɛn pii no esiane SAS Mini-SAS HD a ɛnyɛ den nti (SFF-8643 / . SFF-8654) backplane cable, power supply backplane rail a ɛredi nkogu, anaasɛ drive abien a ɛyɛ dedaw a ɛbɔ sectors bɔne bere koro mu. - Sɛ drive bay LED bi yɛ sum koraa (ɛnyɛ spinning up): Power down server no, unlatch saa drive caddy no, twɛn 30 seconds, firmly resat caddy no na hwɛ SAS backplane power/data cables no, na power san sɔ sɛnea ɛbɛyɛ a physical disks nyinaa bɛhunu ansa na woadi kan ahwɛ import no.

Anamɔn 2: Bue "Amannɔne Nsiesiei Nhwɛso" Screen (Mfa Anifurae Nkɔ Mu Da) .

  1. Boot server no na mia ** .F2 (Nhyehyɛe Nhyehyɛe)** wɔ Dell PowerEdge so 13G / 14G / 15G / 16G server ahorow (anaasɛ ** .Ctrl+R** wɔ mpanyimfo so 11G / 12G BIOS server ahorow a wɔde di dwuma).
  2. Kɔ ** .Device Settings -> [Your Dell PERC / MegaRAID Controller] -> Main Menu -> Configuration Management -> Manage Foreign Configuration -> Preview Foreign Configuration** na ɛyɛ.
  3. Sɛnea ɛbɛyɛ foforo no, sɛ wubetumi abɔ Linux nkwagye USB a ɛte ase anaasɛ wobɛfa SSH/OS so abɔ a, hwɛ preview no dwoodwoo fi ahyɛde kwan no so denam so perccli anaasɛ storcli: 1 .
    # Show all detected Foreign Configurations and whether Virtual Disks are complete:
    perccli /c0/fall show all
    # Or on standard Broadcom MegaRAID:
    storcli /c0/fall show preview
    
  4. Kenkan Preview Screen no Yie:
  5. Hwɛ Virtual Disk (VD 0) status wɔ Amannɔne Preview no mu: So ɛka sɛ ** .Optimal** (draiver nyinaa wɔ hɔ) anaa ** .Degraded** (e.g., draibɛ 6 a ɛwɔ RAID 5 mu no mu 5, anaasɛ draibɛ 8 mu 6 a ɛwɔ RAID 6 mu)?
  6. Mmara: Sɛ Amannɔne Preview no kyerɛ a, wɔde Virtual Disk no bɛba sɛ ** .Optimal** anaa ** .Degraded (Mountable)**, ɛyɛ ahobammɔ sɛ wobɛtoa so akɔ Anammɔn 3.
  7. Gyina Tebea: Sɛ Amannɔne Preview no ka sɛ ** .Offline / . Incomplete** ɛfiri sɛ abien draibɛ da so ara yera firi RAID 5 mu (anaasɛ abiɛsa firi RAID 6 mu), GYINA: MMƆ Import de besi nnɛ. Sɛ wode ba berɛ a draibɛ pii nni hɔ a, ɛbɛfa array a abubu, a wɔrentumi mfa so bɛba anaasɛ ɛbɛma draibɛ a ayera no ayɛ adwuma. Dum server no na hwehwɛ nea enti a honam fam disk ahorow a ayera no ntumi nkɔ soro wɔ backplane no so kan.

Anamɔn 3: Di dwuma Import Foreign Configuration

  1. Sɛ Preview screen no si so dua sɛ Virtual Disk no awie (anaasɛ mountably degraded) wie a, klik ** .Import Foreign Configuration** na si so dua sɛ wobɛka (anaasɛ tu mmirika perccli /c0/fall import / . storcli /c0/fall import).
  2. Wɔ sikɔne 3 kosi 5 mu no, controller no kenkan... DDF metadata fi honam fam disk ahorow no mu, kyerɛw stripe map no kɔ controller no NVRAM mu, dannan drive ahorow no fi Foreign san kɔ ** .Online, na ɛde Virtual Disk no ba ** .Optimal (anaasɛ Degraded, ayɛ krado sɛ wɔbɛkora so ntɛm ara).
  3. San hyɛ server no ase bio; wo dwumadi nhyehyɛe (VMware ESXi, . Windows Server, . Proxmox / Linux) bɛbɔ tẽẽ afi volume a enni dɛm no mu a baiti biako mpo ayera.

Playbook 2: Na sɛ obi klik "Clear Foreign Configuration" DAN (anaasɛ wɔpopa NVRAM + DDF metadata no nyinaa) nso ɛ?

Na sɛ mfiridwumayɛfo kumaa bi ho popoe anɔpa 2:00, na ohui nso ɛ "Foreign configuration found", klike ** .Clear Foreign Configuration** (a wɔpopa DDF RAID metadata off the disks enti drive biara seesei ka sɛ ** .Unconfigured Good**), na wo Virtual Disk no nyinaa ayera?

MMFA wo nsam; wo data da so ara yɛ ɔha biara mu nkyem 100 wɔ mprɛte no so. Sɛ wobɔ so Clear Foreign Configuration popa nketewa no nkutoo DDF configuration sector a ɛwɔ disk ahorow no awiei pɛɛ; ɛnka ɛnyɛ nea ɔde di dwuma no data stripes (LBA 0 ɛdenam wo mpaapaemu no awiei so).

Nanso, adeyɛ a ɛnteɛ biako a wobɛtu bere a wo ho retɔ no no bɛsɛe biribiara. Ɛha na adwumayɛfo data-recovery protocol pɛpɛɛpɛ bere a wɔayi hardware RAID config bi afi hɔ:

Ɔkwan 1 (Ahobammɔ sen biara - 100% Asiane a Ɛwɔ Mfitiase Disk Ho): San Yɛ Ɛkame ayɛ sɛ denam Read-Only HBA + so R-Studio / . UFS Explorer / . mdadm

Professional data recovery engineers da nsɔ "Re-creating the array" nhwɛ wɔ live production disks a ɛwɔ RAID controller BIOS no mu sɛ wonni ɔha biara mu nkyem 100 awerɛhyem wɔ drive nhyehyɛe pɔtee, stripe kɛse (64 KB, . 128 KB, . 256 KB), ne parity rotation ( .Left Asymmetric vs Left Symmetric). 1. Power down server no na yi numbered drives no. 2. Fa drive ahorow no bata adwumayɛbea a wɔde non-RAID SAS Host Bus Adapter (HBA, te sɛ LSI bi 9207-8i / . 9300-8i flashed to IT-Mode Initiator-Target passthrough firmware) enti operating system no hu raw ankorankoro disk biara tẽẽ a RAID BIOS biara ntwitware mu. 3. Bue UFS Explorer RAID Recovery anaa R-Studio Mfiridwumayɛfo: - Nnwinnade abien no nyinaa wɔ Dell PERC, LSI MegaRAID, ne HPE Smart Array Heuristic Stripe-Order Detectors a wɔde scan raw sectors no, ankasa hu drive nhyehyɛe pɔtee, 64 KB / 256 KB stripe block kɛse, parity rotation, ne DDF fi ase offset wɔ akenkan-nko memory mu, na ma wo copy wo nyinaa .vhdx, . .vmdk, SQL databases, ne fael ahorow a efi virtualized array no so tẽẽ kɔ backup NAS a wonkyerɛw baiti biako mpo nkɔ mfitiase SAS draibɛ ahorow no so.

Ɔkwan 2 ("Re-Create Virtual Disk WITH NO INITIALIZATION" Hardware Trick - Bere a Wonim Parameters 100% Nkutoo)

Sɛ wunni bi a IT-Mode HBA na wunim ne ɔha biara mu nkyem 100 ahotoso pɛpɛɛpɛ ansa na wɔayɛ nhyehyɛe a ɛwɔ Dell PERC / LSI MegaRAID nhyehyɛe no mu (fi nea wɔde akɔ amannɔne iDRAC TSR [Tech Support Report] log anaa build krataa: Physical Disk bay nhyehyɛe pɛpɛɛpɛ, RAID level, Stripe Element Size pɛpɛɛpɛ [e.g., 64 KB wɔ PERC H730 so anaasɛ 256 KB wɔ H740P] so, ne Akenkan/Twerɛ Ho Nhyehyɛe): 1. Wɔ MegaRAID / PERC BIOS mfasoɔ no mu no, bue ** .Create New Virtual Disk** na ɛyɛ. 2. Paw RAID Level koro no ara pɛpɛɛpɛ, Physical Disks koro no ara pɛpɛɛpɛ wɔ ntoatoaso koro no ara mu, ne Stripe Size koro no ara pɛpɛɛpɛ (64 KB / . 256 KB). 3. NSƐM-ANAA-OWU NSƐM: Wɔ ** ase .Initialization** dropdown box, hwɛ hu sɛ wɔde ahyɛ ** .NO INITIALIZATION** (Gyae sɛ No / . None deɛ; NEVER paw Fast Initialize anaasɛ Full Initialize). 4. Sɛ wobɔ Virtual Disk wɔ LSI/PERC controller a ** .Initialization = No**, controller no kyerɛw brand-ne foforo DDF metadata header kɔ reserved sectors wɔ disks no awiei bere a gyaw ɔha biara mu nkyem 100 a ɛwɔ ase user data blocks a wɔmfa wɔn nsa nka. 5. Boot kɔ Akenkan Nkutoo Linux Live USB (SystemRescue) kan na tu mmirika lsblk -f anaasɛ fdisk -l: sɛ wo NTFS, . VMFS, anaa XFS mpaepaemu no pue ntɛm ara a wɔn fael nhyehyɛe nkyerɛwde no nni dɛm, stripe geometry no hyia. (Hyɛ no nsow: N mfi ase nhyɛ akyi Parity Consistency Check kosi sɛ woayɛ backup data a ɛho hia nyinaa akyi, efisɛ sɛ woasusuw stripe size a ɛnteɛ anaa drive nhyehyɛe a ɛnteɛ a, Consistency Check bɛkyerɛw data blocks pa a parity bɔne wom.)

Playbook 3: "Multiple Drive Failures" a wodi ho dwuma wɔ RAID 5 / RAID 6 ne "Punctured Stripe" Trap no mu

Na sɛ wo 6-drive RAID 5 array no yɛ ** koraa nso ɛ.Offline / Dead** efisɛ Drive 2 dii nkogu asram abiɛsa a atwam ni (na obiara anhu efisɛ wɔanhyehyɛ email kɔkɔbɔ ahorow) ne Drive 4 fii ase tow Media Mfomso (Bad Blocks / Sense Code 03/11/00) anɔpa yi?

Warning

Nea enti a ɛnsɛ sɛ wutu mmirika anifuraefo DA Force Online + Rebuild bere a drive abien fi RAID 5 mu wɔ mmere ahorow mu: 1. "Stale Drive" Corruption Disaster: Fa no sɛ Drive 2 wui wɔ January 1st, na server no kɔɔ so yɛɛ adwuma wɔ Degraded mode mu nnafua 90 (kyerɛw SQL nkitahodi foforo ɔpepem pii ne VM nsakrae kɔɔ Drives 0, 1, 3, 4, ne 5). Wɔ April 1st no, Drive 4 bɔ sector bɔne bi na ɛtɔ offline, na ɛbɔ array no. Sɛ wokɔ PERC BIOS mu na wo klik "Force Online" wɔ Drive 2 (a awu fi January) so sen sɛ wobɛbɔ Drive 4, anaasɛ wobɔ mmɔden sɛ wobɛsan akyekye denam Drive 2 no nnafua 90 dedaw stale parity no so a, wobɛsan akɔ akyi na woasɛe faelsystem block biara a wɔasesa wɔ nnafua 90 a atwam no mu, na woadan wo NTFS/VMFS volume no ayɛ no Swiss cheese a wontumi mfa so. 2. "RAID Stripe a Wɔabɔ no" (Bad Block Propagating During Rebuild) Trap: Bere a draibɛ a ɛda so te ase wɔ RAID 5 a asɛe mu hyia Akenkan Mfomso a Wontumi Nnya (URE) bere a wɔresan akyekye no, Dell PERC ne LSI MegaRAID controllers yɛ "Punctured Stripe" adwuma: sɛnea ɛbɛyɛ a rebuild no antumi ansɛe koraa no, controller no hyɛ da kyerɛw synthetic "ECC mfomso agyiraehyɛde a wontumi nsiesie" (puncture) kɔ brand-new replacement drive no so wɔ saa LBA pɛpɛɛpɛ no so enti operating system no nim sɛ data yerae wɔ hɔ. Sɛ wɔtow RAID array bi wie a, daakye rebuild biara trɛw puncture no mu, na ɛretu mmirika chkdsk betumi cascade-fail ɛnne no.

Ahobammɔ 4-Anamɔn Nkwagye ma Multi-Drive Dropped RAID Array:

  1. Fa Controller no TTY / Event Log (perccli /c0 show termlog anaasɛ iDRAC TSR) NKYEKYƐMU:
  2. Ansa na wobɛka drive tebea biara no, kenkan controller no mu firmware terminal log (termlog). Hwɛ bere nsɔano pɔtee bere a draibɛ biara dan fii mu no Online to Failed anaasɛ Foreign.
  3. Key Rule: Drive a ɛbaa offline LAST (nnansa yi ara, e.g., simma 10 a atwam ni nnɛ) wɔ 99.999% up-to-date data na ɛyɛ only drive a wobetumi ahyɛ no dwoodwoo asan akɔ intanɛt so (anaasɛ cloned) de ahyɛ volume no so. Drive a ewui adapɛn anaa asram bi a atwam ni (Stale Drive) ɛsɛ sɛ ɛkɔ so out wɔ array no mu.
  4. Clone "Most Recently Dropped" Failing Drive no so wɔ Healthy Blank Disk so Fa Di Dwuma ddrescue (Ɛnam IT-Mode HBA so):
  5. Efisɛ na karka a wogyaee simma 10 a atwam ni no taa gyae esiane nsa kakraa bi a ɛyɛ bɔne a wontumi nkenkan nti (UREs), sɛ wohyɛ saa draibɛ a ɛdi huammɔ wɔ honam fam no wɔ intanɛt so wɔ RAID kaad no mu na wobɔ mmɔden sɛ wobɛsan ayɛ nnɔnhwerew 14 RAID a, ɛbɛma ne ti a wɔakenkan no abɔ na awu koraa wɔ simma 20 akyi.
  6. Mmom, fa saa draibɛ a ɛdi nkoguo no (ne asɔremma draibɛ a aka nyinaa) bata nea ɛnyɛ RAID ho IT-Mode SAS HBA na fa di dwuma ** .GNU ddrescue** ( ) .ddrescue -d -r1 /dev/sdX drive4_clone.img drive4.map) de mfonini 99.9999% ne sectors a wotumi kenkan no gu drive anaa mfonini fael a ɛyɛ apɔwmuden so, skipping yiye wɔ 3 anaa 4 bad sectors no so a ɛnsɛe ti nhyiam no.
  7. Mount Cloned Array no wɔ Read-Only Mode mu na Copy Out Critical Data Ansa na Woasan Akyekye:
  8. Sɛ wɔyɛ draibɛ a adi nkogu no clone wie (anaasɛ sɛ wohyɛ draibɛ a wɔatow agu nnansa yi ara no wɔ intanɛt so wɔ controller BIOS no mu a), fa OS volume no to Read-Only mode mu (mount -o ro,norecovery) na yi wo database ahorow a ɛho hia sen biara no fi hɔ, . .vhdx/ ..vmdk fael, ne akontabuo kyɛfa kɔ abɔnten NAS ntɛm ara.
  9. Sɛnea Wobɛpopa "Punctured Array" Bere a Woayɛ Wo Data Backup akyi:
  10. Bere a Dell PERC anaa LSI MegaRAID Virtual Disk bi ahu amane wɔ Punctured Stripe (wɔakyerɛ wɔ controller log no mu denam Patrol Read can't fix bad block anaasɛ Strip is punctured), swapping drives dodow biara anaasɛ running consistency checks rennyi puncture frankaa no mfi Virtual Disk no so.
  11. Sɛ wobɔ wo fael/VM ɔha biara mu nkyem 100 backup fi punctured array no so wie a: ɛsɛ sɛ wo Delete the Virtual Disk, Full-Initialize a brand-new Virtual Disk wɔ healthy drives so, san yɛ datastore/partition no, na wosan de wo data no fi backup no mu.

Mpɔtam hɔ ntɛmpɛ nɔma ne adwumakuw no nhyehyɛe ahorow

Ɔman / Ɔmantam Emergency & Cyber Nsɛm a Ɛsisii Nnɔmba Enterprise Server OEM Asɛmpatrɛw-Aho hia Mmoa & Data Recovery Gyinapɛn
Aman Nkabom Kuo 911 (Server dan mu ogya / anyinam ahoɔden / halon asiane); CISA (1-844-SAY-CISA) ma infrastructure a ɛho hia / ayaresabea storage outages Dell ProSupport Asɛmpatrɛw a Ɛho Hia (US): 1-800-945-3355 (bisa PERC / Storage Senior Escalation Engineer; fa gu so iDRAC TSR log ansa na woaka Foreign Config) .
HPE PointNext Asɛmpatrɛw a Ɛho Hia (US): 1-800-633-3600
Broadcom / LSI Nnwumakuw Mmoa: 1-800-867-5016
NIST SP 800-88 / SP 800-209 adwumayɛbea akorae ahobammɔ & sanba akwankyerɛ.
Canada 911 (Nneɛma a Egye Ntɛmntɛm); Canada Asoɛe a Ɛhwɛ Intanɛt So Ahobammɔ So (1-833-CYBER-88) Dell ProSupport Canada: Ɔde ne nsa kyerɛɛ ne so. 1-800-847-4096
HPE Nnwumakuw Mmoa Canada: 1-800-633-3600.
Ahemman Nkabom 999 anaa 112 (Ntɛmpɛ); NCSC (ncsc.gov.uk) Dell ProSupport UK: 0800 587 1456 anaasɛ +44 844 444 3844
HPE Nnwumakuw Mmoa UK: 0800 316 1440 anaasɛ +44 207 660 4401.
Europa Aman Nkabom Kuo 112 (Pan-European nɔma a egye ntɛmpɛ) ENISA (European Union Agency for Cybersecurity) & ISO/IEC 27031 Akwankyerɛ a ɛfa Amanneɛbɔ ne Nkitahodi Mfiridwuma a Wɔasiesie wɔn ho ama Adwumayɛ a ɛbɛkɔ so (IRBC).
Australia & New Zealand 000 (AU) / 111 (NZ); ASD ACSC (1300 292 371) Dell ProSupport Australia: Ɔde ne nsa kyerɛɛ ne so. 1800 812 393 / New Zealand: 0800 441 567
HPE Nnwumakuw Mmoa Australia: 13 11 47.

Tip

Ansa na wobɛfrɛ Dell ProSupport, HPE Pointnext, anaa adwumayɛfo RAID data recovery lab no, fa wo Out-of-Band Diagnostic Bundle (iDRAC TSR / . HPE iLO AHS): Bere mpo a wo server no pow sɛ ɛbɛbɔ ne operating system no efisɛ Virtual Disk no yɛ Foreign anaasɛ Offline, wobɛtumi akɔ server no wɛb-based management port (Dell iDRAC anaa HPE iLO) mu afiri laptop biara a ɛwɔ network no so: - Wɔ Dell PowerEdge so (iDRAC 8 / 9 / 10): Kɔ Maintenance -> SupportAssist -> Start a Collection, hwɛ ** .Storage Logs (TTY / TermLog)**, na klik Export Locally so. Ɛno .zip fael no kura mu RAIDControllerLog.txt a ɛkyerɛ milisekɔn pɔtee a drive biara gyaee, pɛpɛɛpɛ DDF metadata GUIDs, nsensanee no kɛse pɔtee (64 KB vs 256 KB), ne sɛ ebia wɔato nsensanee biara mu anaa. - Wɔ HPE ProLiant so (iLO 5 / 6): Kɔ Information -> Active Health System Log na twe nea ɛwɔ .ahs bundle a wɔde ahyɛ mu.

Nsɛm a wɔtaa bisa

Nsonsonoe bɛn na ɛwɔ "Import Foreign Configuration" ne "Clear Foreign Configuration" ntam wɔ Dell PERC / MegaRAID mu?

Susuw asɛm no ho hwɛ `Foreign` flag as the RAID controller asking: "RAID nhyehyeɛ a wɔakyerɛw wɔ saa hard drives yi so no ne me card no mu memory nhyia; emu nea ɛwɔ he na ɛyɛ nokware?" `Import Foreign Configuration` ka kyerɛ controller no sɛ: "Hard drive ahorow no kura nokware no. Kɔpi RAID nhyehyɛe no FI hard drive ahorow no mu KƆ controller card no mu na fa me volume no hyɛ so a wompopa biribiara." Fa di dwuma `Import` bere biara a wobɛsesa RAID kaad a awu, de wo drive nyinaa akɔ server foforo so, anaasɛ wobɛnya power/backplane glitch abɔ wo array a ɛwɔ hɔ dedaw no offline. `Clear Foreign Configuration` ka kyerɛ controller no sɛ: "Popa RAID metadata no fi saa amannɔne drive ahorow no so na dan no ma ɛnyɛ blank." `Unconfigured Good` disks." Fa di dwuma `Clear` NKUTOO bere a wode spare drive biako a wɔde adi dwuma (a wɔatwe afi server dedaw a wɔakɔ pɛnhyen mu) kɔ server a ɛretu mmirika a ɛyɛ papa mu na wopɛ sɛ wopopa saa spare drive biako no metadata dedaw no sɛnea ɛbɛyɛ a ebetumi asan asi sɛ spare a ɛyɛ hyew no nkutoo.

So metumi atwe me SAS/SATA draibɛ nyinaa afi Dell PowerEdge R740 a awu (a motherboard awu) mu na mede ahyɛ Dell PowerEdge R730 anaa R750 a aka no mu de ahyɛ me server no ase ntɛm ara?

Aane. Wɔfrɛ eyi Disk Roaming anaa Array Migration, ne SNIA `DDF` (Disk Data Format) metadata ma ɛyɛ nea wotumi de ho to so wɔ Dell PERC (`H330`, . `H730`, . `H740P`, . `H755`) ne Broadcom MegaRAID controllers, sɛ wudi mmara abiɛsa akyi a: (1) Fa ne bay nɔma pɔtee (`Bay 0` denam `Bay N`) na fa hyɛ server a wɔde besi ananmu no mu wɔ nhyehyɛe koro no ara mu pɛpɛɛpɛ; (2) Hwɛ sɛ server a wɔde besi ananmu no RAID controller firmware no wɔ major version koro no ara anaa foforo mu na ɛboa RAID level a wɔde di dwuma no; na (3) Wɔ boot wɔ replacement server no so no, hyɛn PERC BIOS no mu, hwɛ Foreign Configuration no, na paw `Import Foreign Configuration`.

Dɛn nti na HPE Smart Array kyerɛɛ "1785 - Slot 0 Drive Array Not Configured" wɔ POST, na dɛn nti na wɔremia F2 ɛyɛ owuyare?

Wɔ Hewlett Packard Adwumayɛbea (`HPE`) ProLiant servers a wɔde Smart Array controllers di dwuma, sɛ SAS backplane power cable no ayɛ loose anaasɛ drive shelf bi gye bere tenten dodo sɛ ɛbɛbɔ akɔ soro wɔ boot mu a, controller no bɛda POST nkra te sɛ `1785 - Drive Array Not Configured` anaasɛ `1789 - Disk Drive(s) Respond` akyi no, nneɛma a wubetumi apaw te sɛ `Press F1 to continue` anaasɛ `Press F2 to accept data loss and re-enable logical drive`. Mmia so da `F2` a wonkenkan nsɛm a wɔka kyerɛ no pɛpɛɛpɛ no yiye: wɔ Smart Array firmware awo ntoatoaso pii so no, mia `F2` bere a drives ayera bere tiaa bi no ka kyerɛ controller no sɛ onnye daa loss volume a ayera no ntom anaasɛ ɔnsan mfi ase logical drive metadata no. Mmom, brɛ anyinam ahoɔden no ntɛm ara, siesie SAS nhama a ayɛ hare anaa abɔnten so shelf ahoɔden no sɛnea ɛbɛyɛ a draibɛ no nyinaa bɛbɔ akɔ soro, na san bɔ bio.

So metumi ayi drive biako afi Hardware RAID 5 anaa RAID 10 array mu, de ahyɛ USB-to-SATA dock mu wɔ me Windows PC so, na mayɛ me fael ahorow no ho kɔpi afi so?

Wɔ RAID 5, RAID 6, RAID 0, anaa RAID 10 (a ɛwɔ 4+ drives): Dabi Wɔ striped arrays mu no, fael ankorankoro biara sõ sen stripe kɛse (`64 KB` to `256 KB`) no atwitwa mu ayɛ no ribbons nketenkete a ne kɛse yɛ kilobyte 64 wɔ drive ahorow nyinaa so bere koro mu, na wɔde XOR parity blocks abɔ mu. Sɛ wode RAID 5 draibɛ biako hyɛ Windows PC mu a, ɛbɛma woahu mpaapaemu a wontumi nkenkan, na sɛ wobɔ Windows Disk Management "Initialize Disk (`MBR/GPT`)" popup wɔ saa drive biako no so, wobɛkyerɛw Stripe 0 no sectors a edi kan no so. (Nea ɛka ho nkutoo ne 2-drive RAID 1 Mirror: wɔ 2-drive RAID 1 ahwehwɛ a ɛnyɛ den mu no, disk biara kura linear copy a edi mũ a ɛfa volume no ho na wobetumi de ahyɛ akenkan nkutoo so wɔ SATA/SAS port a ɛnyɛ RAID so, ɛwom sɛ ɛnsɛ sɛ woda so ara bɔ no da a wotumi kyerɛw wɔ ahwehwɛ no akyi sɛ woayɛ nhyehyɛe sɛ wode bɛsan de ahyɛ RAID card no mu a.)

Me RAID 5 rebuild no akɔ so ayɛ 47% na server no yɛ barely responsive. So ɛsɛ sɛ mesan bɔ server no anaasɛ metwa rebuild no mu?

Nea edi kan no, nnyɛ hard-power-off server no sɛ drive dwumadi LED ahorow no da so ara rebɔ a. Sɛ hardware RAID 5 anaa RAID 6 rebuild brɛ ase kɔ crawl (sɛ nhwɛso no, ɛkɔ fam fi 150 MB/sec kɔ 0.5 MB/sec wɔ ɔha biara mu 47), ɛkame ayɛ sɛ bere nyinaa ɛkyerɛ sɛ source drives a ɛda so te ase no mu biako abɔ patch a ɛyɛ mmerɛw/ɛsɛe magnetic sectors na ɔde sikani 8 wɔ sector biara mu reyɛ emu ERC (`Error Recovery Control`) kenkan mmɔden a wɔsan yɛ. Bere a woda so ara wɔ kwan kɔ mounted volume no so wɔ network no so (sɛ ɛyɛ brɛoo mpo a), ntɛm ara kɔpi wo database backups, sikasɛm fael, ne VM configs a ɛho hia sen biara no fi server no so kɔ NAS anaa USB drive a ɛyɛ soronko so mprempren, efisɛ sɛ saa source drive a ɛrepere no gyae koraa ansa na adu ɔha biara mu nkyem 100 a, rebuild no begyae na volume no betumi akɔ offline.

Akwankyerɛ ahorow a ɛfa ho

Fibea ne aban akwankyerɛ

  • Storage Networking Nnwumakuw Fekuw (SNIA): RAID Disk Data Format a wɔtaa yɛ (DDF) Mfiridwuma ho Nkyerɛkyerɛmu (Version 2.0), a ɛma on-disk nhyehyɛe metadata (COD) ne Amannɔne Nsiesiei import wɔ enterprise RAID controllers nyinaa so.
  • Dell Technologies Nnwumayɛbea Mfiridwuma Nwoma: Dell PowerEdge RAID Controller (PERC 9 / 10 / 11 / 12) Ɔdefo Nkyerɛkyerɛmu: Amannɔne Nsiesiei a Wɔhwɛ So, Cache a Wɔakora So (Pinned Cache), ne RAID Ntrɛwmu a Wɔabɔ no Ahorow a Wosiesie (SLN156890 / . 000139251).
  • Broadcom Inc. (kan no na wɔfrɛ no LSI / Avago): MegaRAID SAS Software Dwumadifo Nkyerɛkyerɛmu ne StorCLI Nhwehwɛmu Nhoma (/cX/fall show preview na /cX/fall import Amannɔne Nsiesiei Ahyɛde).
  • Hewlett Packard Enterprise (HPE): HPE Smart Array SR Gen10 / Gen11 Ɔdefoɔ Nkyerɛkyerɛmu ne Ɔhaw Ho Nsiesiei Akwankyerɛ (Nsɛm a Wɔakora So Dwumadibea RIS Metadata ne POST Mfomso Kood ahorow 1785 to 1799).
  • Ɔman Asoɛe a Ɛhwɛ Gyinapɛn ne Mfiridwuma So (NIST): Nwoma Titiriw 800-209 (Ahobanbɔ Akwankyerɛ a ɛfa Nneɛma a Wɔkora So) ne SP 800-88 Rev. 1 (Akwankyerɛ a ɛfa Nsɛm ho Amanneɛbɔ ho Ahotew).