# Disk Management va Partitioning !!! info "Bu darsda" - 🎯 Qattiq disk, bo'lim (partition) va fayl tizimi (filesystem) nima ekanini tushunasiz - 🎯 `lsblk`, `df`, `du` bilan diskingizni ko'rishni o'rganasiz - 🎯 `fdisk` / `parted` bilan bo'lim yaratasiz (va xavfli joylarni bilib olasiz) - 🎯 LVM — moslashuvchan disk boshqaruvi tizimini bilib olasiz - ⏱️ O'qish vaqti: ~18 daqiqa - 🧩 Kerakli bilim: fayl tizimi tuzilishi, terminal asoslari --- ## 🪝 Kvartira, xonalar va javonlar Tasavvur qiling: yangi, bo'sh bir xonali kvartira oldingiz. Unda hech narsa yo'q — na devor, na mebel. Siz birinchi navbatda nima qilasiz? 1. **Devorlar tikasiz** — katta bo'shliqni alohida xonalarga bo'lasiz: yotoqxona, oshxona, mehmonxona. 2. Keyin har bir xonaga **mebel va javon** qo'yasiz — nimani qaerga qo'yishni aniqlaydigan tartib o'rnatasiz. 3. Shundan keyingina xonaga **narsa olib kirib saqlay** boshlaysiz. Qattiq disk ham aynan shunday ishlaydi: - **Disk** — bo'sh kvartira (xom saqlash joyi). - **Partition** (bo'lim) — xonalarga ajratuvchi devorlar. - **Filesystem** (`mkfs`) — har bir xonaga qo'yilgan javon-tartib tizimi. - **Ma'lumot** — javonga terib qo'yilgan narsalar. !!! tip "Oddiy qilib aytganda" Yangi disk sotib oldingiz — bu bo'sh kvartira. Uni ishlatishdan oldin **bo'limlarga ajratasiz** (devor), keyin **fayl tizimi yaratasiz** (javon), shundan keyin fayllar saqlay boshlaysiz. ```mermaid flowchart LR A[Yangi disk
/dev/sdb] --> B[Partition yaratish
fdisk / parted] B --> C[Filesystem yaratish
mkfs.ext4] C --> D[Mount qilish
mount] D --> E[Foydalanish
fayllar saqlash] style A fill:#1e3a8a,stroke:#60a5fa,color:#fff style C fill:#065f46,stroke:#34d399,color:#fff style E fill:#7c3aed,stroke:#a78bfa,color:#fff ``` --- ## Disk turlari va nomlash Terminalda disk ko'rsangiz, `/dev/sda`, `/dev/nvme0n1` kabi nomlar ko'rasiz. Bu nomlar qoidali — bir marta tushunsangiz, boshqa hech qachon chalkashtirmaysiz. ```mermaid flowchart TD root["/dev/ — barcha qurilmalar"] --> sata["SATA/SCSI disklar
/dev/sda, /dev/sdb, /dev/sdc"] root --> nvme["NVMe disklar
/dev/nvme0n1, /dev/nvme1n1"] root --> lvm["LVM logical volume
/dev/mapper/vg-lv"] root --> raid["Software RAID
/dev/md0, /dev/md1"] sata --> parts["/dev/sda1, /dev/sda2
(birinchi diskning bo'limlari)"] nvme --> nvmeparts["/dev/nvme0n1p1, /dev/nvme0n1p2
(NVMe diskning bo'limlari)"] style root fill:#1e3a8a,stroke:#60a5fa,color:#fff ``` !!! tip "Oddiy qilib aytganda" `/dev/sda` — birinchi disk (a = birinchi). `/dev/sdb` — ikkinchi disk (b = ikkinchi). `/dev/sda1` — birinchi diskning birinchi bo'limi. `/dev/nvme0n1p1` — NVMe disk, 0-raqamli, 1-namespace, 1-bo'lim. ```bash # Tizimingizda qanday disklar bor — ko'rish lsblk # Misol chiqish: # NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT # sda 8:0 0 50G 0 disk # ├─sda1 8:1 0 1G 0 part /boot # ├─sda2 8:2 0 20G 0 part / # └─sda3 8:3 0 29G 0 part /home # sdb 8:16 0 100G 0 disk <-- yangi, bo'sh disk ```

🔬 Chuqurroq — disk interfeys turlari (HDD, SSD, NVMe)

**Fizik disk turlari:** | Tur | To'liq nomi | Xususiyat | |-----|-------------|-----------| | HDD | Hard Disk Drive | Aylanadigan disk, sekin (~100–200 MB/s), arzon, katta sig'im | | SATA SSD | Solid State Drive (SATA) | Flash xotira, tez (~500 MB/s), HDD'dan qimmat | | NVMe SSD | Non-Volatile Memory Express | Eng tez (~3000–7000 MB/s), PCIe orqali ulangan | **Tarmoq saqlash (Network Storage):** - **NFS** (Network File System) — tarmoq orqali ulashilgan papka - **iSCSI** — tarmoq orqali ulashilgan blok qurilma (disk kabi ko'rinadi) - **Ceph** — taqsimlangan saqlash tizimi, katta klasterlar uchun **Virtual qurilmalar:** - `/dev/loop0` — faylni disk kabi ko'rsatadigan loopback qurilma - `/dev/mapper/...` — LVM yoki dm-crypt mantiqiy qurilmalari - `/dev/md0` — dasturiy RAID qurilmasi

--- ## Diskni ko'rish buyruqlari ### Nima bor — tezkor ko'rinish ```bash # Barcha block qurilmalar (disk, bo'lim, mount) daraxt ko'rinishda lsblk # Fayl tizimi turi va UUID bilan lsblk -f # O'zingiz tanlagan ustunlar bilan lsblk -o NAME,SIZE,FSTYPE,MOUNTPOINT,UUID ``` ```bash # Har bir mount point qancha joy egallagani (human readable = odam o'qiydigan format) df -h # Misol chiqish: # Filesystem Size Used Avail Use% Mounted on # /dev/sda2 20G 12G 7.0G 63% / # /dev/sda1 976M 150M 760M 17% /boot # /dev/sda3 29G 5.2G 22G 20% /home # Inode ishlatilishi (fayl soni limiti) df -i ``` ```bash # Bitta papkaning hajmi du -sh /var/log # Barcha yuqori darajali papkalar du -sh /* 2>/dev/null # Eng katta papkalarni topish du -sh /var/* 2>/dev/null | sort -rh | head -10 ``` !!! tip "df va du — farqi nima?" `df` — disk bo'limining **umumiy holati** (jami/ishlatilgan/bo'sh). `du` — konkret **papka yoki fayl** qancha joy olayotgani. "Disk to'lib qoldi" xatoligida: avval `df -h` (qaysi bo'lim to'la), keyin `du -sh` (nima to'ldirgan). ### Batafsil disk ma'lumoti ```bash # Barcha disklar va bo'limlarni ro'yxat qilib ko'rsatish sudo fdisk -l # Bitta disk haqida sudo fdisk -l /dev/sda # UUID va fayl tizimi turini ko'rish blkid blkid /dev/sda1 # parted orqali ko'rish sudo parted -l ```

🔬 Chuqurroq — kengaytirilgan disk ma'lumotlari

```bash # Disk apparati haqida batafsil ma'lumot sudo hdparm -I /dev/sda # SMART — diskningning sog'liq holati sudo smartctl -a /dev/sda # Barcha SMART ma'lumot sudo smartctl -H /dev/sda # Sog'liq xulosasi (PASSED/FAILED) sudo smartctl -t short /dev/sda # Qisqa o'z-o'zini test # GPT bo'lim jadvali sudo gdisk -l /dev/sda # I/O statistika — disk qanchalik band iostat iostat -x 1 # Har soniyada yangilanuvchi kengaytirilgan statistika iotop # Qaysi jarayon ko'proq disk ishlatayotgani ```

--- ## 💻 O'zingiz sinab ko'ring — diskni o'qing ```bash # 1. Tizimingizda qanday disklar borligini ko'ring lsblk # 2. Har bir bo'lim qancha joy egallaganini ko'ring df -h # 3. /var papkasi qancha joy olayotganini ko'ring du -sh /var # 4. Disk turlari va UUID'larni ko'ring lsblk -f ``` !!! question "Tekshirib ko'ring" `df -h` va `du -sh /` ni ishga tushirdingiz. `df -h` `/` uchun 8G ishlatilgan deyapti, `du -sh /` esa 7.5G deyapti. Nima uchun farq bor?

Javob — bosib oching

`df` — fayl tizimining **metadata va bloklash tizimi** hisobiga kiritilgan joyni ko'rsatadi. `du` faqat fayllarning haqiqiy hajmini hisoblaydi. Bundan tashqari: o'chirilgan, lekin hali biror jarayon tomonidan ochiq turilgan fayllar `df`da hisoblanadi, `du`da esa ko'rinmaydi. `lsof | grep deleted` bunga sabab bo'layotgan jarayonlarni topadi.

--- ## Partitioning — bo'limlarga ajratish ### MBR va GPT — ikki tizim Bo'lim jadvali disk boshida joylashadi va "qayerdan qayergacha qaysi bo'lim" ni belgilaydi. Ikki eski/yangi standart bor: ```mermaid flowchart LR subgraph MBR["MBR — eski (1983)"] m1["Maks. 2 TB disk"] m2["Maks. 4 ta asosiy bo'lim"] m3["BIOS bilan ishlaydi"] end subgraph GPT["GPT — yangi (2000+)"] g1["2 TB dan katta disklar"] g2["128 tagacha bo'lim"] g3["UEFI bilan ishlaydi"] g4["Zaxira bo'lim jadvali"] end style MBR fill:#7f1d1d,stroke:#f87171,color:#fff style GPT fill:#065f46,stroke:#34d399,color:#fff ``` !!! tip "Oddiy qilib aytganda" Yangi disk — **GPT** ishlatang. MBR faqat eski tizimlar bilan muvofiqlashtirishda kerak bo'ladi. --- ### fdisk — interaktiv bo'lim yaratish !!! danger "Ma'lumotni o'chiruvchi buyruq" `fdisk` bilan bo'lim o'chirish yoki qayta yaratish disknidagi barcha ma'lumotni yo'q qilishi mumkin. Faqat **bo'sh disk yoki maxsus ajratilgan disk** ustida ishlang. Ishonchsiz bo'lsangiz — avval `lsblk` va `fdisk -l` bilan tasdiqlang. ```bash # fdisk ni interaktiv rejimda ishga tushirish # /dev/sdb — bo'sh disk (sda emas!) sudo fdisk /dev/sdb ``` ``` # fdisk ichidagi asosiy buyruqlar: # p — bo'lim jadvalini ko'rsatish (print) # n — yangi bo'lim yaratish (new) # d — bo'lim o'chirish (delete) # t — bo'lim turini o'zgartirish (type) # a — boot flag qo'yish/olib tashlash # w — o'zgarishlarni saqlash va chiqish (write) ← faqat shu yerda o'zgaradi # q — saqlamasdan chiqish (quit) ← xavfsiz chiqish # Amaliy misol — /dev/sdb diskida 10 GB bo'lim yaratish: sudo fdisk /dev/sdb > n # yangi bo'lim > p # primary (asosiy) tur > 1 # bo'lim raqami > Enter # boshlanish sektori (default) > +10G # o'lcham (10 gigabayt) > w # saqlash va chiqish ``` > [!NOTE] > `w` ni bosmaguningizcha hech qanday o'zgarish diskirga yozilmaydi. `q` bilan xavfsiz chiqsa bo'ladi. --- ### gdisk — GPT bo'limlari uchun ```bash # gdisk — GPT bo'lim jadvali bilan ishlash sudo gdisk /dev/sda # gdisk buyruqlari fdisk ga o'xshash: # p — ko'rsatish # n — yangi bo'lim # d — o'chirish # t — tur o'zgartirish # x — expert rejim # w — saqlash # q — saqlamasdan chiqish # GPT afzalliklari: # - 2 TB dan katta disklar uchun # - 4 dan ortiq asosiy bo'lim (128 tagacha) # - Yaxshiroq ma'lumot yaxlitligi # - UEFI bilan muvofiqligi ``` --- ### parted — universal vosita ```bash # Interaktiv rejim sudo parted /dev/sdb # parted ichidagi buyruqlar: # print — bo'lim jadvalini ko'rsatish # mklabel gpt — GPT jadval yaratish # mklabel msdos — MBR jadval yaratish # mkpart — bo'lim yaratish # rm — bo'limni o'chirish # quit — chiqish ``` ```bash # Interaktivsiz (skript ichida ishlatish uchun): # GPT jadval yaratish sudo parted /dev/sdb mklabel gpt # Bo'limlar yaratish (foiz bilan — qulay) sudo parted /dev/sdb mkpart primary ext4 0% 50% # birinchi yarmi sudo parted /dev/sdb mkpart primary ext4 50% 100% # ikkinchi yarmi ```

🔬 Chuqurroq — avtomatik bo'limlash skripti

```bash #!/bin/bash # partition-disk.sh — yangi disk uchun avtomatik bo'limlash DEVICE="/dev/sdb" PARTITION_SCHEME="gpt" # Qurilma tekshiruvi if [ ! -b "$DEVICE" ]; then echo "Xato: $DEVICE blok qurilma emas" exit 1 fi # Ogohlanish — foydalanuvchi tasdiqlashi shart echo "OGOHLANTIRISH: Bu $DEVICE dagi barcha ma'lumotni o'chiradi!" read -p "Davom etasizmi? (yes/no): " confirm if [ "$confirm" != "yes" ]; then echo "Bekor qilindi" exit 0 fi # GPT jadval yaratish sudo parted "$DEVICE" mklabel "$PARTITION_SCHEME" # Boot bo'limi (512 MB) sudo parted "$DEVICE" mkpart primary fat32 1MiB 513MiB sudo parted "$DEVICE" set 1 boot on # Root bo'limi (20 GB) sudo parted "$DEVICE" mkpart primary ext4 513MiB 20.5GiB # Home bo'limi (qolgan joy hammasi) sudo parted "$DEVICE" mkpart primary ext4 20.5GiB 100% # Natijani ko'rsatish echo "Bo'lim jadvali yaratildi:" sudo parted "$DEVICE" print echo "" echo "Keyingi qadamlar:" echo "1. Fayl tizimi yaratish: mkfs.ext4 /dev/sdb2" echo "2. Mount qilish: mount /dev/sdb2 /mnt" echo "3. /etc/fstab ni yangilash (doimiy mount uchun)" ```

--- ## Filesystem yaratish — `mkfs` Bo'lim yaratilgandan keyin uni "formatlash" kerak — fayl tizimi o'rnatiladi. Bu kvartira misolida "javon va tartib o'rnatish" bosqichi. !!! danger "Ma'lumotni o'chiruvchi buyruq" `mkfs` (make filesystem) buyrug'i bo'limdagi barcha mavjud ma'lumotni **to'liq o'chiradi**. Faqat yangi yoki bo'sh bo'limda ishlatiladi. Noto'g'ri qurilmaga ishlatmang. ```bash # ext4 — Linux uchun standart fayl tizimi sudo mkfs.ext4 /dev/sdb1 sudo mkfs.ext4 -L "MyData" /dev/sdb1 # yorliq bilan # XFS — yuqori unumdorlikli fayl tizimi (katta serverlar) sudo mkfs.xfs /dev/sdb1 sudo mkfs.xfs -f /dev/sdb1 # mavjud fayl tizimini majburiy o'chirish # Btrfs — snapshot imkoniyatli zamonaviy fayl tizimi sudo mkfs.btrfs /dev/sdb1 sudo mkfs.btrfs -L "BtrfsVolume" /dev/sdb1 # FAT32 — Windows bilan muvofiqlik (flesh-disk, boot bo'lim) sudo mkfs.vfat -F 32 /dev/sdb1 # Swap bo'limi (virtual RAM) sudo mkswap /dev/sdb2 sudo swapon /dev/sdb2 # faollashtirish ``` !!! tip "Qaysi fayl tizimini tanlash kerak?" - **ext4** — hamma holat uchun yaxshi tanlov. Oddiy, ishonchli, tez. - **XFS** — katta fayllar va yuqori yuk (video, ma'lumot bazasi). Red Hat/CentOS'da default. - **Btrfs** — snapshot, kompressiya kerak bo'lganda. Openwall/SUSE serverlarida. - **FAT32** — Windows bilan almashinuv, UEFI boot bo'limi.

🔬 Chuqurroq — mkfs parametrlari va tekshiruv

```bash # ext4 qo'shimcha parametrlar sudo mkfs.ext4 -b 4096 /dev/sdb1 # blok o'lchami (odatda 4096) sudo mkfs.ext4 -m 1 /dev/sdb1 # zahira joy foizi (default 5%, serverda 1% yaxshi) sudo mkfs.ext4 -j /dev/sdb1 # journal bilan (default, xavfsiz) sudo mkfs.ext4 -O ^has_journal /dev/sdb1 # journalsiz (tezroq, lekin crash'da xavfli) # XFS parametrlar sudo mkfs.xfs -b size=4096 /dev/sdb1 # blok o'lchami sudo mkfs.xfs -s size=512 /dev/sdb1 # sektor o'lchami sudo mkfs.xfs -d agcount=4 /dev/sdb1 # allocation groups soni (parallel I/O uchun) # Fayl tizimini yaratilgandan keyin tekshirish sudo fsck.ext4 /dev/sdb1 # ext4 tekshiruv sudo xfs_check /dev/sdb1 # XFS tekshiruv ```

--- ## LVM — Moslashuvchan Disk Boshqaruvi ### Muammo: qattiq bo'lingan bo'limlar Tasavvur qiling: `/home` uchun 50 GB ajratdingiz. Bir yil o'tib, u to'lib qoldi, lekin `/var` bo'limida 30 GB bo'sh turibdi. Oddiy bo'limlashda buni qayta taqsimlash — ma'lumotlarni o'chirish va formatdan o'tkazish demak. **LVM** (Logical Volume Manager — Mantiqiy Hajm Menejeri) bu muammoni hal qiladi: bo'limlarni real vaqtda o'lchamini o'zgartirish, bir necha diskni birlashtirib ishlatish, snapshot olish mumkin. ```mermaid flowchart TB subgraph PVs["Physical Volumes (PV) — Fizik qurilmalar"] pv1["/dev/sdb1
50 GB"] pv2["/dev/sdc1
50 GB"] end subgraph VG["Volume Group (VG) — Umumiy havuz"] pool["mydata VG
100 GB umumiy joy"] end subgraph LVs["Logical Volumes (LV) — Virtual bo'limlar"] lv1["lv_web
20 GB"] lv2["lv_db
40 GB"] lv3["Erkin joy
40 GB"] end pv1 --> pool pv2 --> pool pool --> lv1 pool --> lv2 pool --> lv3 style VG fill:#1e3a8a,stroke:#60a5fa,color:#fff style lv3 fill:#374151,stroke:#9ca3af,color:#9ca3af ``` !!! tip "Oddiy qilib aytganda" PV — alohida idishlar (disklar). VG — ularni quyib bir katta idish hosil qilish. LV — o'sha katta idishdan kerakli miqdorda olish. Istalgan vaqt LV o'lchamini o'zgartirish mumkin — disk to'kib almashtirishsiz. ### LVM o'rnatish va sozlash ```bash # LVM vositasini o'rnatish sudo apt install lvm2 # Ubuntu/Debian sudo yum install lvm2 # CentOS/RHEL ``` ```bash # 1-QADAM: Physical Volume (PV) yaratish sudo pvcreate /dev/sdb1 sudo pvcreate /dev/sdb2 # PV'larni ko'rish sudo pvdisplay sudo pvs # qisqa format ``` ```bash # 2-QADAM: Volume Group (VG) yaratish # Bir yoki bir necha PV'dan bitta VG hosil qilish sudo vgcreate mydata /dev/sdb1 /dev/sdb2 # VG'larni ko'rish sudo vgdisplay sudo vgs # qisqa format ``` ```bash # 3-QADAM: Logical Volume (LV) yaratish sudo lvcreate -L 10G -n volume1 mydata # aniq 10 GB sudo lvcreate -l 50%VG -n volume2 mydata # VG'ning 50% # LV'larni ko'rish sudo lvdisplay sudo lvs # qisqa format ``` ```bash # 4-QADAM: Fayl tizimi yaratish va mount qilish sudo mkfs.ext4 /dev/mydata/volume1 sudo mkdir /mnt/volume1 sudo mount /dev/mydata/volume1 /mnt/volume1 ``` ### LVM boshqaruvi ```bash # Volume Group kengaytirish — yangi disk qo'shish sudo pvcreate /dev/sdb3 sudo vgextend mydata /dev/sdb3 # Logical Volume kengaytirish — hajm qo'shish sudo lvextend -L +5G /dev/mydata/volume1 # 5 GB qo'shish sudo lvextend -l +100%FREE /dev/mydata/volume1 # bo'sh joyning hammasini olish # Fayl tizimini kengaytirish (LV kengaytirilgandan keyin shart!) sudo resize2fs /dev/mydata/volume1 # ext2/3/4 uchun sudo xfs_growfs /mnt/volume1 # XFS uchun (mount qilingan holda) ```

🔬 Chuqurroq — LVM qisqartirish, o'chirish, snapshot

**LV qisqartirish (faqat ext2/3/4, XFS qisqarmaydi):** !!! danger "Ma'lumotni o'chiruvchi amal" LV qisqartirish xato bajarilsa ma'lumot yo'qoladi. Oldin **zaxira** (backup) oling! ```bash # Avval umount qilish shart sudo umount /dev/mydata/volume1 # Fayl tizimini tekshirish sudo e2fsck -f /dev/mydata/volume1 # Avval fayl tizimini kichraytirish sudo resize2fs /dev/mydata/volume1 8G # Keyin LV ni kichraytirish sudo lvreduce -L 8G /dev/mydata/volume1 ``` **LV va VG o'chirish:** ```bash sudo umount /dev/mydata/volume1 sudo lvremove /dev/mydata/volume1 # LV o'chirish sudo vgremove mydata # VG o'chirish (avval barcha LV o'chirilgan bo'lishi kerak) sudo pvremove /dev/sdb1 # PV o'chirish ``` **LVM Snapshot — zaxira nuqtasi:** ```bash # Snapshot yaratish (shu paytdagi holat) sudo lvcreate -L 1G -s -n volume1-snap /dev/mydata/volume1 # Snapshotni mount qilib ko'rish sudo mkdir /mnt/snapshot sudo mount /dev/mydata/volume1-snap /mnt/snapshot # Snapshotdan tiklash (asl LV shu holatga qaytadi) sudo umount /dev/mydata/volume1 sudo umount /dev/mydata/volume1-snap sudo lvconvert --merge /dev/mydata/volume1-snap # Snapshotni o'chirish (tiklamasdan) sudo umount /dev/mydata/volume1-snap sudo lvremove /dev/mydata/volume1-snap ```

--- ## RAID — Ishonchlilik va Unumdorlik

🔬 Chuqurroq — Software RAID (mdadm)

RAID (Redundant Array of Independent Disks) bir necha diskni birlashtirib, yo tezlikni, yo ishonchlilikni oshiradi. Uy serverlarida ham, korporativ infratuzilmada ham keng qo'llanadi. **RAID darajalari:** | RAID | Nomi | Tavsif | Minimum disk | |------|------|---------|--------------| | 0 | Striping | Tezlik, lekin zahirasiz — bitta disk xato = hammasi yo'q | 2 | | 1 | Mirroring | Bir xil nusxa — bitta disk xato bo'lsa ikkinchisi ishlaydi | 2 | | 5 | Striping + parity | Tezlik + bitta disk xatosiga chidamlilik | 3 | | 6 | Double parity | Ikkita disk xatosiga chidamlilik | 4 | | 10 | RAID 1+0 | Mirroring + Striping | 4 | ```bash # mdadm o'rnatish sudo apt install mdadm # Ubuntu/Debian sudo yum install mdadm # CentOS/RHEL # RAID 1 (mirror) yaratish sudo mdadm --create --verbose /dev/md0 \ --level=1 \ --raid-devices=2 \ /dev/sdb1 /dev/sdc1 # RAID 5 yaratish sudo mdadm --create --verbose /dev/md1 \ --level=5 \ --raid-devices=3 \ /dev/sdb1 /dev/sdc1 /dev/sdd1 # RAID holatini ko'rish cat /proc/mdstat sudo mdadm --detail /dev/md0 # RAID konfiguratsiyasini saqlash sudo mdadm --detail --scan | sudo tee -a /etc/mdadm/mdadm.conf sudo update-initramfs -u ``` ```bash # RAID boshqaruvi sudo mdadm --add /dev/md0 /dev/sdb3 # zahira disk qo'shish sudo mdadm --fail /dev/md0 /dev/sdb2 # nosoz diskni belgilash sudo mdadm --remove /dev/md0 /dev/sdb2 # olib tashlash sudo mdadm --add /dev/md0 /dev/sdb4 # yangi disk qo'shish (avtomatik tiklanadi) # RAID to'xtatish sudo umount /dev/md0 sudo mdadm --stop /dev/md0 # RAID butunlay o'chirish sudo mdadm --zero-superblock /dev/sdb1 /dev/sdc1 ```

--- ## Disk holati va xizmat ko'rsatish ### Sog'liq monitoringi ```bash # SMART — disk o'z diagnostikasi (Self-Monitoring Analysis and Reporting Technology) sudo smartctl -H /dev/sda # Sog'liq xulosasi: PASSED yoki FAILED sudo smartctl -a /dev/sda # Barcha SMART ma'lumot sudo smartctl -t short /dev/sda # Qisqa o'z-o'zini test (2 daqiqa) sudo smartctl -t long /dev/sda # To'liq test (bir necha soat) sudo smartctl -l selftest /dev/sda # Test natijalari # Disk harorati sudo hddtemp /dev/sda ```

🔬 Chuqurroq — unumdorlik testi va disk tozalash skripti

**Unumdorlik testi:** ```bash # hdparm — oddiy o'qish testi sudo hdparm -t /dev/sda # Buffered disk o'qish sudo hdparm -T /dev/sda # Cache o'qish # dd bilan benchmark sudo dd if=/dev/zero of=/tmp/testfile bs=1M count=1024 conv=fdatasync sudo dd if=/tmp/testfile of=/dev/null bs=1M count=1024 rm /tmp/testfile # fio — professional benchmark (o'rnatish kerak) fio --name=random-read \ --ioengine=posixaio \ --rw=randread \ --bs=4k \ --numjobs=1 \ --size=1g \ --iodepth=1 \ --runtime=30 \ --time_based ``` **Disk tozalash va xizmat skripti:** ```bash #!/bin/bash # disk-maintenance.sh — muntazam disk xizmat ko'rsatish LOG_FILE="/var/log/disk-maintenance.log" log() { echo "[$(date '+%Y-%m-%d %H:%M:%S')] $1" | tee -a "$LOG_FILE" } # Disk to'lganligini tekshirish check_disk_space() { log "=== Disk holati ===" df -h | awk 'NR>1 && /^\/dev/ { # Foiz raqamini olish gsub(/%/,"",$5) if ($5+0 > 90) print "KRITIK: " $6 " " $5 "% to'la" else if ($5+0 > 80) print "OGOHLANTIRISH: " $6 " " $5 "% to'la" }' | while read line; do log "$line"; done } # Katta fayllarni topish find_large_files() { log "=== 1 GB dan katta fayllar ===" find / -type f -size +1G 2>/dev/null \ -exec ls -lh {} \; | head -20 | while read f; do log "$f" done } # Vaqtinchalik fayllarni tozalash clean_temp() { log "=== Vaqtinchalik fayllarni tozalash ===" n=$(find /tmp -type f -mtime +7 -delete -print 2>/dev/null | wc -l) log "/tmp dan $n ta eski fayl o'chirildi" n=$(find /var/tmp -type f -mtime +30 -delete -print 2>/dev/null | wc -l) log "/var/tmp dan $n ta eski fayl o'chirildi" if command -v apt >/dev/null; then apt autoclean >> "$LOG_FILE" 2>&1 log "APT kesh tozalandi" fi journalctl --vacuum-time=30d >> "$LOG_FILE" 2>&1 log "30 kundan eski journal loglari o'chirildi" } # Asosiy bajarish log "Disk xizmat ko'rsatish boshlandi..." check_disk_space find_large_files clean_temp log "Disk xizmat ko'rsatish yakunlandi" ```

--- ## Filesystem xatolarini tuzatish

🔬 Chuqurroq — tiklash va zaxira

### Fayl tizimini tekshirish va tuzatish !!! warning "Tez-tez uchraydigan xato" `fsck` faqat **umount qilingan** bo'limda ishlaydi. Agar boot bo'limida xato bo'lsa — live USB orqali yuklab, o'sha yerda bajariladi. ```bash # Avval umount qilish shart sudo umount /dev/sdb1 # ext2/3/4 tekshirish va tuzatish sudo fsck.ext4 -f /dev/sdb1 # majburiy tekshiruv sudo fsck.ext4 -y /dev/sdb1 # avtomatik tuzatish (barcha savollarga "ha" deydi) # XFS tekshirish va tuzatish sudo xfs_check /dev/sdb1 # faqat tekshirish sudo xfs_repair /dev/sdb1 # tuzatish # Btrfs sudo btrfs check /dev/sdb1 sudo btrfs check --repair /dev/sdb1 ``` ### Ma'lumotlarni tiklash ```bash # O'chirilgan fayllarni tiklash vositalari sudo apt install testdisk photorec sudo photorec /dev/sdb1 # fayl tiklash sudo testdisk /dev/sdb # bo'lim tiklash # Shikastlangan diskdan nusxa olish sudo apt install ddrescue sudo ddrescue /dev/sdb1 /recovery/sdb1.img /recovery/sdb1.log # dd orqali (ddrescue yo'q bo'lsa) sudo dd if=/dev/sdb of=/recovery/sdb.img bs=1M conv=noerror,sync # Tasvirni mount qilish sudo mount -o loop /recovery/sdb1.img /mnt/recovery ``` ### Bo'lim jadvalini zaxiralash ```bash # Bo'lim jadvalini faylga saqlash sudo sfdisk -d /dev/sda > partition-backup.txt # Tiklash (DIQQAT: diskni to'liq qayta yozadi) sudo sfdisk /dev/sda < partition-backup.txt ```

--- ## 💻 O'zingiz sinab ko'ring — amaliy mashq ### Mashq 1: Disk holatini tekshiring ```bash # 1. Tizimingizning disk tuzilmasini ko'ring lsblk -f # 2. Qaysi bo'lim qancha joy egallaganini ko'ring df -h # 3. Eng ko'p joy olayotgan papkalarni toping du -sh /var/* 2>/dev/null | sort -rh | head -5 # 4. Bo'lim batafsil ma'lumotini ko'ring sudo fdisk -l ``` ### Mashq 2 (VM yoki sinov disk bilan) !!! danger "Faqat sinov muhitida bajarish" Quyidagi buyruqlar real ma'lumotni o'chiradi. Faqat virtual mashinada yoki maxsus ajratilgan sinov diskida bajaring. Asl serverda bajarish qat'iyan man etiladi. ```bash # Yangi bo'sh disk qo'shilganda (masalan, VMda /dev/sdb): # 1. Diskni ko'rish sudo fdisk -l /dev/sdb # 2. Bo'lim yaratish sudo fdisk /dev/sdb # n → p → 1 → Enter → +5G → w # 3. Fayl tizimi yaratish sudo mkfs.ext4 /dev/sdb1 # 4. Mount qilish sudo mkdir /mnt/test_disk sudo mount /dev/sdb1 /mnt/test_disk # 5. Tekshirish df -h /mnt/test_disk ``` !!! question "Tekshirib ko'ring" `df -h` da `/dev/sdb1 5.0G 24M 4.7G 1% /mnt/test_disk` ko'rdingiz. 5 GB disk yaratdingiz, lekin 4.7 GB bo'sh ko'rsatmoqda — 300 MB qayerga ketdi?

Javob — bosib oching

ext4 fayl tizimi odatda umumiy joyning **5%** ni tizim zahirasi sifatida ajratadi. Bu administrator (`root`) ning disk to'lib qolganda ham tizimga kirib muammolashtirishi uchun zarur. `mkfs.ext4 -m 1 /dev/sdb1` bilan zahirani 1% ga tushirsa bo'ladi — bu server disklari uchun yaxshi amaliyot (`root` uchun 5 GB zahira kerak emas).

--- ## Tez-tez uchraydigan xatolar !!! warning "Tez-tez uchraydigan xato: noto'g'ri qurilmani formatlash" `mkfs.ext4 /dev/sda` — bu tizim diskingizni formatlaydi! `/dev/sdb`, `/dev/sdc` kabi qurilmalarni `lsblk` bilan avval tasdiqlang. Ko'p hollarda `sda` — asosiy disk, unga tegmang. !!! warning "Tez-tez uchraydigan xato: resize2fs dan oldin lvextend" LVM'da LV kengaytirilgandan keyin fayl tizimini ham kengaytirish kerak. `lvextend` bitdi — `resize2fs` (ext4) yoki `xfs_growfs` (XFS) ishga tushiring. Aks holda yangi joy ko'rinmaydi. !!! warning "Tez-tez uchraydigan xato: df bilan du farqi" `df -h` disk to'lib qolganini ko'rsatdi, lekin `du -sh /*` deyarli bo'sh deyapti. Sabab: o'chirilgan, lekin hali ochiq fayl bor. `lsof | grep deleted` bilan topib, tegishli jarayonni qayta ishga tushiring. !!! warning "Tez-tez uchraydigan xato: mount qilingan qurilmada fsck" `fsck` faqat umount qilingan bo'limda ishlaydi. Agar root bo'limda xato bo'lsa, live USB orqali yuklab bajarish kerak. --- ## 🎒 Xulosa — eng muhimi Agar bu darsdan faqat beshta narsani eslab qolsangiz: 1. **Disk → Partition → Filesystem → Mount** — bu to'rtta qadam har doim shu tartibda bajariladi. O'tkazib bo'lmaydi. 2. **`lsblk`** — tizimingizda qanday disklar borligini ko'rish. **`df -h`** — har bir bo'lim qancha joy egallagani. **`du -sh`** — konkret papka hajmi. 3. **`fdisk`/`parted`** — bo'lim yaratish. **`mkfs`** — fayl tizimi o'rnatish. Ikkalasi ham ma'lumotni o'chiradi — faqat bo'sh diskda ishlatiladi. 4. **LVM** — dinamik bo'lim boshqaruvi. Disk to'lib qolganda qayta formatlashsiz kengaytirish imkonini beradi. 5. **GPT** — 2 TB dan katta disklar va UEFI uchun. MBR — eski tizimlar uchun. Yangi disklarda doim GPT. !!! success "Keyingi qadam" Endi disklarni boshqarishni bildingiz. Keyingi darsda **Process Management** — Linux'da jarayonlarni ko'rish, boshqarish va avtomatlashtirish bilan tanishamiz. `ps`, `top`, `kill`, `systemd` — bular sizni kutmoqda!