Mugga wax-soo-saarka Docker ama Kubernetes waa la tirtiray: samatabbixin docker system prune --volumes, overlay2 Inodes, iyo etcd Sawir qaade¶

ordaya docker compose down -v, docker system prune -a --volumes, docker volume rm, ama kubectl delete pvc on tab terminal qaldan ee wax soo saarka waa mid ka mid ah siyaabaha ugu dhaqsiyaha badan si ay u tirtirto nool PostgreSQL, MySQL, MariaDB, Redis, MongoDB, ama mugga database SQLite. Si ka duwan sawirada mashiinka farsamada ee daruuriga ah ee u guura 7-maalmood qashinka dib u warshadaynta, darawalka mugga deegaanka ee Docker (/var/lib/docker/volumes/<name>/_data) iyo Kubernetes local-path / hostPath Bixiyeyaashu waxay isla markiiba soo saaraan POSIX caadiga ah unlink() iyo rmdir() wuxuu si toos ah uga soo horjeedaa nidaamka faylalka martida loo yahay ee Linux.ext4, XFS, ama Btrfs). Qoraal kasta oo ku dhaca martigeliyaha Linux ka dib marka amarku dhamaado waxay dib u qori kartaa kaydka xogta ee la sii daayay ee diskka. Haddii saxanka serverkaagu uu buuxsamay boqolkiiba 100 (0 bytes free) ka hor intaan cidina cararin docker system prune --volumes argagax, tasho diskka serverka 100% buuxa, xogta WAL log bloat, iyo soo kabashada eber; faylalka Linux ee biraha qaawan ee tirtirka ka baxsan weelasha, eeg shil-rm -rf ama Shift-Delete on ext4, XFS, APFS, iyo NTFS: dib-u-ururinta akhriska-kaliya iyo xardho-inode iyo mashiinka farsamada ee VMDK, VHDX, iyo qcow2 silsiladda sawir-qaadista musuqmaasuqa iyo samatabbixinta hypervisor. Haddii wax-soo-saar la'aantani ay saamayso daryeelka bukaanka isbitaalka, dirista degdegga ah, ama nidaamka badbaadada utility dadweynaha, wac lambarkaaga xaaladda degdegga ah (911 Maraykanka iyo Kanada, 112 guud ahaan Midowga Yurub, 999 gudaha Boqortooyada Midowday (UK) ama hubi hagahayaga nambarada degdega ah.

Hadda samee tan¶

1. Ha ordi docker compose up -d, systemctl restart docker, ama dib u habayn Kubernetes pod: Dib u samaynta weelka ama dib u bilaabida Docker daemon waxa ay qortaa megabyte cusub overlay2 Xogta badan, weelka JSON, iyo faylalka xog ururinta ee madhan ee dhowaan la bilaabay (initdb) si toos ah uga sarreeya saxda ah ext4 ama XFS blocks halka codkaaga la tirtiray uu fadhiyey ilbiriqsi ka hor.
2. ** Hubi 5 ilbiriqsi haddii habka weelka (postgres, mysqld, redis-server, node, python) ayaa wali ku socota xusuusta (ps aux iyo sudo lsof +L1***: Haddii qof uu tirtiro faylalka gudaha mugga korka ah ama uu ka gooyay mugga qarsoodiga ah iyada oo weelku weli socdo, Linux kernel wuxuu hayaa ** 100 boqolkiiba faylasha xogta aan la xidhin ee noolaanshaha saxanka *** ilaa inta uu ku jiro aqoonsiga habka martigeliyaha weelka (PID) waxa uu haystaa tilmaame fayl furan (FD). Waxaad nuqul ka samayn kartaa faylashaas si toos ah /proc/<PID>/fd/ ama /proc/<PID>/root/ iyadoo xogta eber ay lumineyso ka hor inta uusan weelka joogsan.
3. **Haddii weelka mar hore la joojiyay oo mugga la tirtiray (docker compose down -v ama docker system prune --volumes), qaboojiye diskka martida ayaa isla markiiba qoraya ***: Jooji daemon-ka Docker (sudo systemctl stop docker.socket docker.service containerd) markaa weelasha asalka ah iyo jeegaga caafimaadku waxay joojiyaan qorista logyada /var/lib/docker, ka dibna ama dib u fuul /var/lib/docker (ama xididka / qayb) akhri-kaliya (sudo mount -o remount,ro /ama qaado sawir mug leh oo ah heerka xannibaadda isla markiiba (AWS EBS Snapshot, GCP Persistent Disk Snapshot, DigitalOcean Volume Snapshot) intaanay wax taaban.
4. ** Haddii aad ku sugan tahay Kubernetes (kubectl delete pvc ama kubectl delete namespace), hubi Volume Joogtada ah (PV) ReclaimPolicy (kubectl get pv) isla markiiba**:
5. Haddii ay PV lahaa persistentVolumeReclaimPolicy: Retain, xogtaada lama tirtirin ***: the PV si fudud loogu dhaqaaqay Released status iyo Cloud diskEBS, Azure Disk, Ceph RBD, NFS) 100% waa sax.
6. Xitaa haddii kubectl delete pvc ayaa ku dheggan Terminating, Kubernetes ** Dhammaystirayaasha (kubernetes.io/pvc-protection*** Ka ilaali darawalka CSI inuu tirtiro kaydinta hoose ilaa inta Pod kasta uu wali tixraacayo PVC. Ha tirtirin Pod-ka ilaa aad nuqul ka sameysato xogta Pod-ka socda.
7. Sawir qaybta qaybta hoose ee hoose (dd / ddrescue in ka badan SSH) si disk dibadeed ah ama mugga daruur gaar ah ka hor inta aadan ordin ext4magic, extundelete, ama PhotoRec: Waligaa ha ku shubin fayl-faleyaasha ku qoraya faylalka wax soo saarka isla qaybta diskka ee aad ka soo kabanayso /var/lib/docker/volumes.
8. ** Wixii la musuqmaasuqay ama si lama filaan ah loo tirtiray Kubernetes cluster state (kubectl delete crd / namespace), joogso kube-apiserver oo soo celi kaagii u dambeeyay etcd sawir (etcdctl snapshot restore***: Haddii kooxdu ay sheegto etcd waa la tirtiray iyadoo mugga kaydinta daruuraha hoose ay weli jiraan, soo celinta etcd soo kabsado dhammaan PV, PVC, Secretiyo ConfigMap walxaha isku mar.

Geedka go'aanka: mugga Docker ee la tirtiray, overlay2 lakabka, ama Kubernetes PVC kala soocidda¶

Isticmaal jaantuskan si aad u dooratid dariiqa soo kabashada eber ee ugu dhaqsaha badan iyadoo ku saleysan in habka weelku uu wali nool yahay, haddii mugga uu taageeray saxanka deegaanka ama kaydinta daruuraha CSI, iyo nidaamka faylalka martigeliyay /var/lib/docker:

flowchart TD
    Start["Xogta Wax-soo-saarka ee maqan Kadib Docker ama Kubernetes Tirtir Taliska"] --> WhichEnv{"Waa kuwee Madal & Taliskii Khasaaraha dhaliyay?"}

    WhichEnv -->|"Docker: Koonteenar Wali Soconaya"| ProcRescue["Samatabbixinta Sharaxayaasha Faylka aan Xidhiidhka Lahayn ee Furan iyada oo loo sii marayo sudo lsof +L1 iyo /proc/PID/fd/N ama /proc/PID/root/"]

    WhichEnv -->|"Docker: weel la joojiyay & mugga la tirtiray (hoos u koob -v / prune --volumes)"| StopDocker["Isla markiiba Orod: sudo systemctl stop docker.socket docker weel ku jira"]

    WhichEnv -->|"Kubernetes: kubectl tirtir pvc/pv/statefulset/namespace"| K8sCheck{"Boodhku wali ma shaqaynayaa mise PVC-ku ma ku dheggan yahay joojinta?"}

    K8sCheck -->|"Haa: Pod wali wuu socdaa (Finalizer xannibay CSI tirtiridda)"| K8sExec["Ha tirtirin Pod -> Kubectl exec / pg_dump / tar dusha sare ee stdout isla markiiba ku dheji diskka maxalliga ah"]
    K8sCheck -->|"Maya: Pod Gone -> Hubi kubectl heli pv"| ReclaimCheck{"Maxay ahayd Siyaasadda Dib-u-soo-celinta PV?"}

    ReclaimCheck -->|"Haysashada (Xaalada PV = La sii daayay)"| ClearClaimRef["100%."]
    ReclaimCheck -->|"Tirtir (Cloud EBS / PD / Longhorn / Waddada Maxalliga ah)"| CloudOrLocal{"Mugga CSI ee Cloud ama Disk-ga Martigelinta Maxaliga ah?"}

    CloudOrLocal -->|"Cloud CSI (AWS EBS / GCP / Azure / Ceph)"| CloudSnap["Hubi Cloud Console Dib-u-warshadaynta Bin/Automated EBS Snapshots / Ceph RBD Trash (rbd qashinka ls)"]
    CloudOrLocal -->|"Disk Node Local (dariiqa-maxali ah / Waddada martida loo yahay)"| StopDocker

    StopDocker --> RemountRO["Remount Host Partition Read-Kaliya (Moun -o remount,ro) AMA Stream Block Sawirka SSH"]
    RemountRO --> HostFS{"Waa maxay Nidaamka Faylasha Martigeliyay /var/lib/docker?"}
    HostFS -->|"ext4"| Ext4Rec["Isticmaal debugfs / ext4magic Journal Inode Transaction Rollback (-timestamp)"]
    HostFS -->|"XFS/Btrfs"| CarveRec["Adeegso Xididada Btrfs / xfs_irecover / Saxeexa Saxeexa Gaarka ah ee SQLite, Postgres Heap, ama InnoDB .ibd"]

Marxaladda 1: Sida Docker iyo Kubernetes ay dhab ahaantii u tirtiraan mugga diskka¶

Kahor intaadan socodsiin amar kasta oo soo kabasho ah, fahan meesha xogtaada jidh ahaan ugu noolayd Linux host-ga iyo waxa amarka tirtirku taabtay:

Wajiga 2: Ka badbaadinta faylasha xogta furan ee weelka weli socda (/proc/<pid>/fd/)¶

Sug: ** weli weelku hadda ma socdaa?** Inta badan, injineerku waxa uu wadaa qoraalka socdaalka codsiga, shaqada cron nadiifinta, ama buug-gacmeedka rm -rf /var/lib/postgresql/data/* ama rm /app/data/production.sqlite3 gudaha weel socda, ama tirtira galka bind-mount host ee hoose /opt/app/data halka weelka laftiisa (postgres, mysqld, mariadbd, redis-server, mongod, gunicorn, node) waa wali waa socdaa oo socda.

Linux, faylka lagama tirtiro saxanka marka rm ama unlink() ayaa lagu magacaabaa. Xirmooyinka saxanka iyo inodesku waxay sii jiraan ** 100 boqolkiiba waa la qoondeeyey waana la akhriyi karaa ** illaa habka ugu dambeeya ee haya sharraxaha faylka furan (FD) in inode uu xidho ama ka baxo.

Talaabada 1: Soo hel Aqoonsiga Habraaca martida loo yahayPID) habka weelka¶

On marinka martigeliyaha Linux (ma aha gudaha weelka), hel martigeliyaha PID ee xogtaada ama habka codsiga:

sudo lsof +L1 | grep -E "postgres|mysqld|mariadbd|sqlite|redis|mongo|node|python|java"

COMMAND     PID   USER   FD   TYPE DEVICE  SIZE/OFF NLINK    NODE NAME
mysqld    18492    999   14u   REG  259,1 482344960     0 1849201 /var/lib/mysql/app_prod/users.ibd (deleted)
mysqld    18492    999   15u   REG  259,1 125829120     0 1849202 /var/lib/mysql/app_prod/orders.ibd (deleted)
mysqld    18492    999   16u   REG  259,1  79691776     0 1849190 /var/lib/mysql/ibdata1 (deleted)
python3   22104   1000    4u   REG  259,1  89128960     0 3102944 /data/production.sqlite3 (deleted)

Talaabada 2: Si toos ah uga soo guuri faylasha xogta xogta ee aan la xidhin /proc/<PID>/fd/¶

**ha joojin ama dib ha u bilaabin weelka. Inta hawshu wali socoto, koobi ka bixi sharraxayaasha faylka cayriin si toos ah gal ammaan ah oo ku yaal martidaada:

sudo mkdir -p /root/proc_emergency_rescue

# Example 1: Rescuing an unlinked SQLite3 database (PID 22104, FD 4)
sudo cp /proc/22104/fd/4 /root/proc_emergency_rescue/production.sqlite3
# Check if PID 22104 also holds the -wal file open (e.g. FD 5) and copy it too:
sudo cp /proc/22104/fd/5 /root/proc_emergency_rescue/production.sqlite3-wal

# Example 2: Automatically rescuing ALL open (deleted) MySQL/Postgres files for PID 18492
for fd in /proc/18492/fd/*; do
    target=$(sudo readlink "$fd" 2>/dev/null)
    if echo "$target" | grep -q "(deleted)"; then
        fd_num=$(basename "$fd")
        clean_name=$(echo "$target" | sed 's/ (deleted)//; s/\//_/g')
        echo "Rescuing FD $fd_num -> $clean_name"
        sudo cp "$fd" "/root/proc_emergency_rescue/${fd_num}${clean_name}"
    fi
done

Talaabada 3: Xitaa ka sii fiican: Samee qashin toosan oo macquul ah (pg_dump / mysqldump / .backup) ka hor intaan weelku xirin FD-yada¶

Hadii postgres ama mysqld ayaa wali socda iyada oo faylasha xogta ay u furan yihiin xusuusta iyo /proc/<PID>/fd/, isku day inaad ku shubto qashin toos ah oo toos ah godka unix ama xidhiidhka TCP ee jira marka hore:

# If the database engine can still read its open file descriptors:
docker exec <container_name> pg_dumpall -U postgres > /root/proc_emergency_rescue/live_dump.sql
# Or for MySQL:
docker exec <container_name> mysqldump -u root -p --all-databases --single-transaction > /root/proc_emergency_rescue/live_mysql.sql

Marxaladda 3: Soo kabashada mugga Docker ee la tirtiray ext4 ama XFS ka dib markii weelku istaagay¶

Ka warran haddaad orod docker compose down -v ama docker system prune --volumes, sidaas darteed weelka waa horeyba u joogsaday iyo dhammaan /proc/<pid>/fd/ Sharaxayaashu ma xiran yihiin? Hadda waxaad ku jirtaa tartan ka dhan ah qoraalka asalka ah ee Linux martigeliyaha.

Tallaabada 1: Ku qabooji nidaamka faylka martida loo yahay Linux 10 ilbiriqsi gudahood¶

1. Jooji Docker, weel, kubelet, iyo cron isla markiiba markaa weel, jeeg caafimaad, ama rotator log uma qorayo byte kale /var/lib/docker:

   sudo systemctl stop docker.socket docker.service containerd.service cron.service
   

2. ** Hubi qalabka xannibaya iyo nidaamka faylalka /var/lib/docker**:

   df -Th /var/lib/docker
   

   Filesystem     Type  Size  Used Avail Use% Mounted on
   /dev/nvme0n1p1 ext4  100G   42G   58G  42% /
   

3. ** Dib ugu dheji qaybta Read-Kaliya (ama ku shub sawirka qaybinta cayriin ee SSH mashiin kale)**:
4. Hadii /var/lib/docker waxay ku jirtaa qaybteeda gaarka ah ama mugga LVM (tusaale ahaan /dev/nvme0n1p2 ama /dev/mapper/vg0-docker), ku celi akhri-kaliya isla markiiba:

   sudo mount -o remount,ro /var/lib/docker
   

5. Hadii /var/lib/docker waa xididka / qayb (/Isku day inaad ku qasbi qaybta xididka akhri-kaliya adoo isticmaalaya kernel Magic SysRq kicinta (echo u | sudo tee /proc/sysrq-trigger) ama sudo mount -o remount,ro /.
6. **Haddii aad ku jirto daruur VPS ah (AWS EC2, GCP, DigitalOcean, Hetzner, Linode): Isla markiiba u tag konsole webka daruurta oo guji Create Snapshot mugga xannibaadda ee server-ka (ama ku shub qalabka xannibaadda cayriin ee SSH server labaad: sudo dd if=/dev/nvme0n1p1 bs=4M status=progress | gzip -1 | ssh user@backup-server "cat > /mnt/storage/docker_host_raw.img.gz").

Tallaabada 2: Hubi in xogtaadu ay weli ku fadhido mug kale oo aan la jarin ama overlay2 lakabka¶

Kahor intaadan xardhamin baloogyada saxanka ee aan la qoondayn, xaqiiji in xogtaada aan lagu kaydin buurta xidha ama mug si kale loo magacaabay oo badbaaday:

# Search the entire /var/lib/docker tree read-only for any surviving database files:
sudo find /var/lib/docker -name "PG_VERSION" -o -name "ibdata1" -o -name "*.sqlite*" -o -name "*.db" -o -name "dump.rdb" -o -name "WiredTiger" 2>/dev/null

Talaabada 3: Soo kabashada tusaha mugga Docker ee aan la xidhin ext4 isticmaalaya ext4 Joornaal (ext4magic iyo debugfs)¶

Goorma docker volume rm tirtirtaa /var/lib/docker/volumes/<vol_name>/_data ku ah ext4 nidaamka faylka, ext4 eber ka saarayso tilmaamayaasha tooska ah (i_block / xadka xidid geed) ee miiska inode live. Si kastaba ha ahaatee, **ext4 joornaal (inode <8>, sida caadiga ah 128 MB ku 1024 MB Wareegtada giraangiraha ah docker compose down -v ayaa lagu fuliyay.

+-------------------------------------------------------------------------+
| HOW ext4magic RECOVERS A DELETED DOCKER VOLUME FROM THE EXT4 JOURNAL    |
|                                                                         |
| Live Inode Table on Disk (After docker volume rm):                      |
|   Inode #918402 (users.ibd): [Size = 0, Extent Pointers = ZEROED]       |
|                                                                         |
| Ext4 Journal Ring Buffer (Inode <8> on Disk):                           |
|   Transaction #41902 (10 mins ago):                                     |
|   Inode #918402 (users.ibd): [Size = 482MB, Extents = Blocks 9000..9999]|
|                                                                         |
| -> ext4magic reads the old Inode Copy from Journal Transaction #41902   |
|    and extracts Blocks 9000..9999 directly to your rescue drive.        |
+-------------------------------------------------------------------------+

Orod ext4magic lid ku ah qaybtaada akhris-kaliya (ama .img koobiga sawir-qaadista ee server-ka labaad), oo tilmaamaya waqti-shabad (timestamp)-a) laga bilaabo 5 daqiiqo ka hor inaad orodday docker compose down -v ama docker system prune --volumes:

# Install ext4magic on your rescue machine (never on the affected partition)
sudo apt-get install -y ext4magic e2fsprogs

# Dump the ext4 journal from the frozen partition / image to a file in RAM (/dev/shm)
sudo debugfs -R "dump <8> /dev/shm/ext4_journal.copy" /dev/nvme0n1p1

# Recover all files deleted under var/lib/docker/volumes after (Current Time - 15 minutes)
sudo mkdir -p /mnt/external_rescue/docker_recovered
sudo ext4magic /dev/nvme0n1p1 \
  -j /dev/shm/ext4_journal.copy \
  -f var/lib/docker/volumes \
  -a $(date -d "-15 minutes" +%s) \
  -r -d /mnt/external_rescue/docker_recovered

Tallaabada 4: Xarida SQLite cayriin, PostgreSQL, ama MySQL InnoDB faylasha marka joornaalku duubo (XFS ama ext4)¶

Haddii martigeliyaha isticmaalo XFS ama haddii ay ext4 joornaal hore ugu duudduubay tilmaamayaasha inode, isticmaal saxeexa iyo qaab-dhismeedka xardho ka soo horjeeda sawirka saxanka cayriin (docker_host_raw.img):

1. ** Loogu talagalay SQLite3 Databases (.sqlite, .db - tusaale, Vaultwarden, Nextcloud, Kaaliyaha Guriga, Gitea, Grafana)**: Xog kasta oo SQLite3 ah waxay ku bilaabataa madax 16-byte ASCII ah oo aan la garan karin oo ah byte 0 ee boggeeda ugu horreeya ee 4 kilobyte: 53 51 4c 69 74 65 20 66 6f 72 6d 61 74 20 33 00 ("SQLite format 3\0"), oo kaydisa cabbirka bogga xogta saxda ah (Bytes 16..17, sida caadiga ah 4096) iyo guud ahaan cabbirka xogta ee boggaga (Bytes 28..31, big-endian 32-bit integer) isla gudaha madaxa isla.
2. Orod PhotoRec leh [File Opt] -> [x] sqlite karti, ama si toos ah ugu sawir qaybta grep -ab "SQLite format 3" si aad u heshid kaydka saxda ah ee kaydka saxda ah ee faylkaaga SQLite ee tirtiray oo aad ku soo saarto dd:

   # Find the exact byte offsets of every "SQLite format 3" header on the partition:
   sudo grep -ab "SQLite format 3" /dev/nvme0n1p1
   

   Marka la soo saaro, orod sqlite3 carved.db ".recover" | sqlite3 repaired.db in dib loo dhiso bog kasta oo caleen B-geed ah.

1. ** Loogu talagalay MySQL / MariaDB InnoDB (.ibd faylasha miiska)**: InnoDB kasta .ibd shaxdu waxay ka kooban tahay taxane 16 KB (16,384 byte) bogag ka kooban FIL_PAGE_INDEX (0x45BF) madaxa iyo miiska gaarka ah space_id. Isticmaal Percona / TwinDB undrop-for-innodb (stream_parser + c_parser* Qalabka wax lagu sawiro /dev/nvme0n1p1 dhammaan 16 KB InnoDB index bogag oo saf kasta si toos ah ugu tuur faylasha TSV/SQL si dib loogu soo dejiyo.
2. ** Loogu talagalay Faylasha Heap PostgreSQL ($PGDATA/base/<db_oid>/<relfilenode>***: PostgreSQL waxay u kaydisaa miis kasta si xidhiidhsan 8 KB (8,192 byte) bogag taallo oo ka kooban PageHeaderData (pd_lower, pd_upper, pd_special). Isticmaal pg_filedump ama pglazarus / pdu si loo sawiro gobollada diskooga aan la qoondayn 8 KB bogag u dhigma qaabka tuple tiirka miiskaaga (HeapTupleHeaderData) oo dib u dhis safafka.

Wajiga 4: Kubernetes PVC, PV, iyo etcd Soo kabashada Degdegga ah ee Sawirka¶

Kubernetes (EKS, GKE, AKS, k3s, RKE2, kubeadm, OpenShift), si lama filaan ah u tirtiraya Dacwada Joogtada ah (PersistentVolumeClaim)kubectl delete pvc), StatefulSet, ama meel magac oo dhan ah (kubectl delete ns prod) waxa uu raacayaa qaab dhismeedka aad uga duwan kan Docker-ka taagan.

Muuqaalka A: PVC-ga ayaa muujinaya Terminating (The Finalizer Lifesaver)¶

Orod:

kubectl get pvc -n <namespace>

# Stream a compressed tar backup or pg_dump out of the still-running pod immediately:
kubectl exec -n <namespace> <pod-name> -- pg_dumpall -U postgres > ./rescued_k8s_db.sql
# Or tar an entire mounted volume directory out over stdout:
kubectl exec -n <namespace> <pod-name> -- tar czf - /var/lib/postgresql/data > ./rescued_pvc_files.tar.gz

Dhacdada B: PVC-ga waa la tirtiray, iyo kubectl get pv wuxuu muujinayaa Persistent Volume in Released gobolka (ReclaimPolicy: Retain)¶

Orod:

kubectl get pv

Sida loo sameeyo a Released PV Available mar labaad 10 ilbiriqsi gudahood xogta eber 1. Wax ka beddel shayga joogtada ah oo tirtir spec.claimRef xannibid:

kubectl patch pv <your-pv-name> --type json -p '[{"op": "remove", "path": "/spec/claimRef"}]'

Dhacdada C: Soo celinta gobolka Kubernetes ee la tirtiray oo ka yimid a etcd sawir (etcdctl snapshot restore)¶

Ka warran hadduu qof ordo kubectl delete namespace prod ama helm uninstall oo tirtiray 200 oo muuqaal oo Kubernetes ah, siraha, ConfigMaps, iyo qeexitaannada PV ee is-maamul kubeadm, k3s, ama RKE2 diyaarada gacanta ku haysa?

1. ** Hel kii ugu dambeeyay etcd sawir-qaade *** (sida caadiga ah gudaha /var/lib/rancher/k3s/server/db/snapshots/, /var/lib/rancher/rke2/server/db/snapshots/, ama jadwalkaaga /var/backups/etcd/snapshot.db).
2. ** Xaqiiji daacadnimada sawirka ka hor intaadan taaban kooxda tooska ah ***:

   ETCDCTL_API=3 etcdctl snapshot status /path/to/snapshot.db --write-out=table
   

3. Jooji kube-apiserver iyo etcd ku dul duleelka dayaarada xakamaynta:
4. Daaran kubeadm (boorayaasha taagan): dhaqaaq /etc/kubernetes/manifests/kube-apiserver.yaml iyo /etc/kubernetes/manifests/etcd.yaml ku meel gaar ah si /root/manifests_hold/ sidaas kubelet u joojiya labada weel si nadiif ah, ka dibna dib u soo celi /var/lib/etcd ku /var/lib/etcd.bak.
5. Daaran k3s: Orod sudo systemctl stop k3s iyo sudo k3s server --cluster-reset --cluster-reset-restore-path=/path/to/snapshot.
6. ** Soo celi etcd tusaha xogta daaran kubeadm**:

   sudo ETCDCTL_API=3 etcdctl snapshot restore /path/to/snapshot.db \
     --data-dir=/var/lib/etcd \
     --name=<control-plane-node-name> \
     --initial-cluster=<node-name>=https://<node-ip>:2380 \
     --initial-advertise-peer-urls=https://<node-ip>:2380
   

7. Dhaqaaq etcd.yaml iyo kube-apiserver.yaml dib ugu noqo /etc/kubernetes/manifests/. 30 ilbiriqsi gudahood, kube-apiserver dib ugu xidhidhiyaha la soo celiyay etcd dukaanka iyo meel kasta oo la tirtiro Magaca, Sirta, ConfigMap, iyo Shayga Joogtada ah ee Volume ayaa dib u soo baxaya.

Marxaladda 5: Adkeynta Docker iyo Kubernetes mugga lidka ku ah tirtiridda shilalka ah¶

Marka xogtaada wax-soo-saarka la soo celiyo, dhaqangeli afartan dhisme ee ilaalinta si hal qori oo CLI ah uusan waligiis mar dambe u tirtirin mugga nool:

• ** Beddel wax soo saar kasta Kubernetes StorageClass ka reclaimPolicy: Delete ku reclaimPolicy: Retain**: Sida caadiga ah, fasalada Kaydinta daruuraha ee ku dhex dhisan (gp2, gp3, standard-rwo) ku haboon reclaimPolicy: Delete, taasoo la micno ah in PVC la tirtiro waxay burburisaa saxanka daruuriga ah. Samee a StorageClass leh reclaimPolicy: Retain (ama ku dheji PV-yada jira kubectl patch pv <pv-name> -p '{"spec":{"persistentVolumeReclaimPolicy":"Retain"}}') sidaas darteed tirtirida PVC ama Namespace weligeed ma tirtirto mugga xannibaadda hoose.
• Dari AWS EBS Recycle Bin/GCP Jadwalada Sawir-qaadista/Velero oo wata CSI Snapshots: Awood-siinta AWS EBS Recycle Bin waxay haynaysaa mugga EBS ee tirtiray 7 ilaa 30 maalmood xataa haddii darawalka CSI uu soo waco DeleteVolume.
• ** Waligaa ha isticmaalin docker system prune --volumes qoraalada nadiifinta otomaatiga ah ama CI / CD runners ***: Haddii weelka xogta wax soo saarka si kooban u shil galo (Exited (1)) wakhtiga saxda ah shaqo cron ah ayaa socota docker system prune -a --volumes -f, Docker waxay tixgalinaysaa muggiisa qarsoodiga ah "aan la isticmaalin" oo si joogto ah u tirtira. Xarig --volumes laga bilaabo dhammaan qoraallada nadiifinta.
• **Isticmaal xirmooyinka xidhidhiyaha cad/srv/prod_data/postgres:/var/lib/postgresql/data) ama xog ururin dibadda ah oo la maareeyey oo loogu talagalay culaysyada shaqo ee wax-soo-saar dawladeed ***: Ka duwan /var/lib/docker/volumes/, marti geliso hagaha xidhxidhan ee dibadda /var/lib/docker/ waa ** waligood lama taaban *** by docker compose down -v, docker volume rm, ama docker system prune --volumes.

Lambarada degdega ah ee gobolka iyo borotokoolka wakaalada¶

Marka weel ama tirtiridda kaydinta Kubernetes ay saamayso hawlgallada caafimaad ee isbitaalka, diritaanka badbaadada dadweynaha, ama xogta shakhsiyeed ee nidaamsan, isla markiiba way kacdaa:

• Mareykanka iyo Kanada: garaac 911 haddii kooxdu weel guuldarraysato ay curyaamiso la socodka bukaan-socodka, nidaamyada EHR, ama diritaanka degdegga ah. U soo sheeg kaabayaasha weelka xaasidnimada ah tirtirista ama shilalka internetka CISA (1-844-Say-CISA / cisa.gov/report ama Xarunta Amniga Internetka ee Kanada1-833-CYBER-88).
• United Kingdom: Dial 999 khataraha nolosha ama kaabayaasha NHS ee muhiimka ah. Sida hoos timaada UK GDPR, burburinta joogtada ah ama luminta helitaanka xogta shakhsiyeed iyada oo aan la helin kayd la soo kaban karo waxay u baahan tahay qiimaynta jebinta 72-saac ee ogeysiiska ICO (0303 123 1113).
• Midowga Yurub: garaac 112 dhammaan dalalka xubnaha ka ah Midowga Yurub. Hay'adaha hoos yimaada NIS2 Dardaaranka waa in ay digniin hore u gudbiyaan CSIRT-ga qaranka 24 saac gudahooda ka dib adeeg la'aan daran, iyo luminta xogta shakhsi ahaaneed ee joogtada ah ee kiciya ** GDPR Qodobka 33** warbixinta 72 saacadood gudahood.
• ** Australia iyo New Zealand ***: garaac 000 gudaha Australia ama 111 gudaha New Zealand; u soo sheeg dhacdooyinka kaabayaasha muhiimka ah ACSC (1300 CYBER1).
• **India iyo Asia-Pacific ***: garaac 112 gudaha Hindiya si wax looga qabto xaalad degdeg ah; u soo sheeg shilalka amniga daruuraha ama weelka CERT-In (incident@cert-in.org.in) gudaha daaqadda qasabka ah ee 6-saac ah.

Su'aalaha soo noqnoqda¶

sameeya docker compose down -v Tirtir xirmooyinka xirmooyinka martida loo yahay (./data:/var/lib/mysql) Mise Mugyada Docker ee la magacaabay iyo kuwa aan la garanayn?¶

`docker compose down -v` KALIYA tirtirtaa mugga la magacaabay ee lagu dhawaaqay `volumes:` qaybta `docker-compose.yml` iyo mugga qarsoon ee ku dheggan weelasha; waa marna tirtirtaa faylalka ku dhex jira fuudhka xidhidhiyaha (sida `- /opt/mysql_data:/var/lib/mysql` ama `- ./postgres_data:/var/lib/postgresql/data`). Haddii faylkaaga qoraalka ahi uu isticmaalo dariiqa martida loo yahay ee ka bilaabmaya `/` ama `./`, Ka hubi galkaas martigeliyaha hadda: feylashaadu weli boqolkiiba 100 waa ay saxsan yihiin.

Maxaa yeelay docker system prune --volumes tirtir mugga xog ururintayda markii aan joojiyay weelka 2 daqiiqo oo keliya si aan u cusbooneysiiyo doorsoomayaasha deegaanka?¶

Si taxadar leh u sug kala soocida `docker stop` iyo `docker rm`: haddii weel kaliya yahay `Stopped / Exited` (ka muuqda gudaha `docker ps -a`), `docker system prune --volumes` sameeyo maaha Tirtir muggiisa sababtoo ah weelka la joojiyay ayaa weli tixraacaya iyaga. Si kastaba ha ahaatee, haddii aad orday `docker rm ` (si aad dib ugu samayso weelka calamo cusub) ama haddii aad gudubtay `docker system prune -a --volumes` (kaas oo marka hore tirtiraya dhammaan weelasha la joojiyay iyo markaas wuxuu tirtiraa dhammaan mugga aan ku xidhnayn weel kasta), mugga wuxuu noqday "jiidasho" millise seconds weelka la joojiyay waa la tirtiray oo isla markiiba waa la tirtiray.

Ma soo ceshan karaa mugga Docker ee tirtiray gudaha Docker Desktop ee Mac (Apple Silicon / Intelama Windows (WSL2)¶

Haa, haddii aad isla markiiba ka baxdo Docker Desktop si mashiinka farsamada uu u joojiyo socodsiinta `fstrim` ka dhanka ah saxanka farsamada. On macOS, Docker wuxuu ku kaydiyaa dhammaan mugga Linux gudaha `-/Library/Containers/com.docker.docker/Data/vms/0/data/Docker.raw`; on Windows WSL2, gudaha `%LocalAppData%\Docker\wsl\disk\docker_data.vhdx`. Nuqul ka samee `.raw` ama `.vhdx` isla markiiba u fayl geli SSD dibadeed, ku dheji akhri-kaliya Linux (adoo isticmaalaya `losetup -Pr` waayo `.raw` ama `qemu-nbd -r` waayo `.vhdx`), oo orod `ext4magic` ama `PhotoRec` lid ku ah farsamada casriga ah `ext4` qaybsanaan gudaha.

Maxaa dhacaya haddii aan ku tirtiro Kubernetes StatefulSet kubectl delete statefulset postgres? Taasi ma tirtirtay PVC-yadaydii sidoo kale?¶

Maya. Sida caadiga ah ee Kubernetes (haddii `persistentVolumeClaimRetentionPolicy.whenDeleted: Delete` waxaa si cad loogu habeeyey Kubernetes 1.27+), oo socda `kubectl delete statefulset` waxay tirtirtaa maamulaha StatefulSet iyo Pods-keeda, oo dhan ka tago `PersistentVolumeClaims` (`data-postgres-0`, `data-postgres-1`) iyo xogtooda hoose 100% waa sax. Kaliya dib u codso `StatefulSet` YAML oo leh magac isku mid ah iyo `volumeClaimTemplates`, iyo Pods-ka cusub ayaa si toos ah dib ugu dhejin doona kuwa jira `data-postgres-0` PVC-yada

Sideen u geli karaa faylasha gudaha Docker overlay2 hagaha marka weelku diido inuu bilaabo (CrashLoopBackOff / Exited (1))¶

Ha tirtirin weelka shil galay. Ilaa inta weelka la joojiyay uu weli ku jiro `docker ps -a`, lakabkeeda la qori karo iyo kor u qaadista mugga ayaa si buuxda ugu xidhan saxanka. Waxaad si toos ah uga koobi kartaa galalka weelka la joojiyay adiga oo isticmaalaya `docker cp :/var/lib/postgresql/data /root/rescued_data`, ama soo hel martigeliyaha saxda ah `overlay2` dariiqa orod `docker inspect --format '{{.GraphDriver.Data.UpperDir}}'` iyo akhrinta buuggaas si toos ah sida `root`.

Ka warran haddii qof tirtiro Kubernetes Namespace (kubectl delete ns prod) iyo dhammaan PVC-yada waxay lahaayeen reclaimPolicy: Delete AWS EBS ama GCP?¶

Haddii darawalka CSI ee daruuraha (`ebs.csi.aws.com` ama `pd.csi.storage.gke.io`) horay loo fuliyay `DeleteVolume` sababtoo ah `reclaimPolicy` ahaa `Delete`, hubi sadexda shabaqyada badbaadada heerka daruuriga ah isla markiiba: (1) hubi ** AWS Dib-u-warshadeynta Bin ee EBS Volumes** ee AWS EC2 console (haddii xeerka sii hayntu uu firfircoonaa koontada/tagagga, mugga EBS ee la tirtiray wuxuu ku fadhiyaa Dib-u-warshadeynta Binka waxaana lagu soo celin karaa 1 guji); (2) ka eeg **AWS Backup/DLM/GCP Jadwalka Sawirka Sawirka** sawirka mugga ee ugu dambeeyay; iyo (3) hubi haddii xogtaadu ahayd mid sii daynaysay WAL/binlogs (`WAL-G`, `pgBackRest`, `Litestream`) baaldi S3/GCS.

Tilmaamaha la xidhiidha¶

• Saxanka server-ka 100% buuxa (0 bytes bilaash ah): xogta WAL log bloat, B-geed kharribmay, iyo soo kabashada eber
• Shift rm-rf ama `Shift+Delete' ee Linux, macOS, ama Windows: isla markiiba dib u soo kicida akhrinta kaliya iyo xardhitaanka
• Mashiinka Virtual VMDK, VHDX, iyo qcow2 silsiladda sawir-qaadista musuqmaasuqa iyo samatabbixinta hypervisor
• Git shil ah ee furayaasha AWS, faylka .env, ama furaha gaarka ah ee SSH ee dadweynaha GitHub repo
• Xiisadda martigelinta shabakada, socdaalka, iyo soo celinta kaabta
• Kombiyuutarka kaydinta, 3-2-1 xeeladda, iyo soo kabashada musiibada
• Nambarada degdega ah wadan ahaan

Ilaha iyo hanuuninta rasmiga ah¶

• Dukumentiyada rasmiga ah ee Kuberneteskubernetes.io / Aasaaska Xisaabinta Dhaladka Cloud): Waxyaabaha Kaydinta ee Ilaalinta Isticmaalkakubernetes.io/pvc-protection Dhamaystirka), Soo Celinta Codad Joogta ah (Retain vs. Delete), iyo Hawlgelinta etcd Kooxaha Kubernetes (etcdctl snapshot restore)
• Docker Inc.docs.docker.com): * Ku maamul xogta gudaha Docker: Volumes, Bind Mounts, overlay2 Dhismaha Darawalka Kaydinta, iyo Tixraaca Prune*.
• Ururka Kernel Linuxkernel.org): * The /proc Nidaamka faylka (procfs Fur Soo kabashada Faylka Sharaxaada iyada oo loo marayo /proc/<pid>/fd/) iyo Ext4 Journal Specification*.
• ext4magic Dukumentiga Mashruuca (ext4magic.sourceforge.net): * Inode Forensic iyo Joornaalka Dib u soo kabashada Ganacsiga ee Ext3/Ext4 Filesystems*.
• Machadka Heerarka iyo Tignoolajiyada Qaranka (NIST(Mareykanka): Daabacaada Gaarka ah 800-190: Codsiga Hagaha Amniga Konteenarka & Soo Kabashada Dhacdada.