This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision | ||
|
storage:svc_recover [2025/11/04 16:34] manu |
storage:svc_recover [2025/11/10 11:07] (current) manu [Recover a SVC / v3k / v5k / v7k] |
||
|---|---|---|---|
| Line 1: | Line 1: | ||
| - | ==== Recover a SVC / v3k / v5k / v7k ==== | + | ===== Recover a SVC / v3k / v5k / v7k ===== |
| + | |||
| + | FIXME Do not use this procedure without IBM support | ||
| + | |||
| + | Node 1 starting and LED code 578: | ||
| - | démarrage node 1 | ||
| - | 578: | ||
| On the active node, connect to service IP https://sevice_IP/service | On the active node, connect to service IP https://sevice_IP/service | ||
| --> select the first node | --> select the first node | ||
| + | --> Place Node in Service State | ||
| --> manage system | --> manage system | ||
| --> remove system data on both nodes, and reboot if not candidate | --> remove system data on both nodes, and reboot if not candidate | ||
| + | This command removes configuration, cache, and location data from the specified node. | ||
| + | |||
| + | !!DANGER!! satask removeclusterdatacanister -node <node_canister_id> \\ | ||
| + | !!DANGER!! satask leavecluster -force | ||
| **Ensure that all nodes in the system of nodes to be recovered display a status of candidate.** | **Ensure that all nodes in the system of nodes to be recovered display a status of candidate.** | ||
| reboot node1 | reboot node1 | ||
| - | recover system | + | |
| - | prepare for recovery (5-15 min) | + | ===== Recover system ===== |
| + | |||
| + | prepare for recovery (5-15 min) service web interface or cmd: satask t3recovery -prepare | ||
| <code> | <code> | ||
| Line 33: | Line 42: | ||
| </code> | </code> | ||
| - | --> recover (allow popup on Web browser) | + | --> recover (allow popup on Web browser) or cmd: satask t3recovery -execute |
| <code> | <code> | ||
| IBM_2076:v7000-c1:superuser>sainfo lscmdstatus | IBM_2076:v7000-c1:superuser>sainfo lscmdstatus | ||
| Line 123: | Line 132: | ||
| You have to wait the end of recovery to have a full operational SVC. | You have to wait the end of recovery to have a full operational SVC. | ||
| - | ==== Post procedure ==== | + | ===== Post procedure ===== |
| - | === Recovering from offline VDisks using the CLI === | + | ==== Recovering from offline VDisks using the CLI ==== |
| Any volumes that are offline and are not thin-provisioned (or compressed) volumes are offline because of the loss of write-cache data during the event that led all node canisters to lose their cluster state. **Any data lost from the write-cache cannot be recovered.** These volumes might need additional recovery steps after the volume is brought back online. | Any volumes that are offline and are not thin-provisioned (or compressed) volumes are offline because of the loss of write-cache data during the event that led all node canisters to lose their cluster state. **Any data lost from the write-cache cannot be recovered.** These volumes might need additional recovery steps after the volume is brought back online. | ||
| * Delete all IBM FlashCopy function mappings and Metro Mirror or Global Mirror relationships that use the offline volumes. | * Delete all IBM FlashCopy function mappings and Metro Mirror or Global Mirror relationships that use the offline volumes. | ||
| - | * Run the **recovervdisk** or **recovervdiskbysystem** command. (This will only bring the volume back online so that you can attempt to deal with the data loss.) | + | * If there are corrupted volumes in a pool, and the volumes are thin-provisioned or compressed, use: |
| + | **repairsevdiskcopy** //_vdisk_name_// | ||
| + | * If the volume is not a thin-provisioned or compressed volume, and it is outside of a data reduction pool, use: | ||
| + | **recovervdiskbysystem** | ||
| + | * Else try: | ||
| + | **recovervdisk** //_vdisk_name_// | ||
| + | |||
| + | (This will only bring the volume back online so that you can attempt to deal with the data loss.) | ||