Isolate or repair local disks - Elastic Compute Service - Alibaba Cloud Documentation Center

Isolate or repair a damaged local disk on an ECS instance via the console or API after receiving a disk damage event.

Important

Isolating or repairing a disk causes unrecoverable data loss on the damaged local disk. Back up your data in advance.

During disk isolation or repair, you can redeploy the instance with local disks to restore full disk usage, but all disk data will be lost.

Procedure overview

Modify configuration files: For Linux instances, log on to the instance, modify the /etc/fstab file, and unmount the damaged local disk to prevent read/write operations on the wrong disk during replacement.
Isolate a damaged disk: Isolate the damaged local disk. The instance continues to operate normally after isolation.
Replace the disk: Alibaba Cloud replaces the damaged disk.
Restore the disk: Log on to the ECS instance, modify the /etc/fstab file, and mount the new local disk.

Note

Only big data instance families d1, d2, and d3, and local SSD instance family i4 support online disk isolation and replacement. Local SSD instance families i3 and i2 support isolation but not replacement. The actual steps displayed on the page depend on the extent of disk damage.

Procedure

Modify configuration files (Linux only)

Skip this step if your instance does not run Linux. For Linux instances, modify the /etc/fstab file and unmount the damaged local disk before isolation:

Modify configuration files

Connect to the Linux instance.

If local disks are configured for automatic mounting at startup in /etc/fstab, add the nofail and barrier parameters for all local disks to prevent mount failures from blocking instance startup.

Before modification:

UUID=cf4572d0-****-*** /path/to/mountpoint fstype defaults 0 0

After modification:

UUID=cf4572d0-****-*** /path/to/mountpoint fstype defaults,nofail,barrier=0 0 0

Parameter	Description
UUID	The UUID of the local data disk. Run the `blkid` command to query it.
/path/to/mountpoint	The mount point of the local disk. Run the `blkid` command to query it.
fstype	The file system type. Run the `blkid` command to query it.
barrier	Whether to enable the write barrier. 0 means disabled, and 1 means enabled.
nofail	Mount failure does not block instance startup.

Unmount the damaged local disk.
Important
If you do not unmount the damaged disk, the device name may change after repair, causing applications to read or write to the wrong disk.
```
umount /path/to/mountpoint
```

Isolate a damaged disk

ECS console

Visit ECS Console-Events.
In the left-side navigation pane of the Event page, click Local Disk-based Instance Events > Local Disk Damaged Events. Click Repair in the Actions column of the target event.
In the Configurations Modification step, confirm that you have modified the configuration file, and click Next.

The page indicates that you need to modify the /etc/fstab configuration file in the ECS instance operating system and unmount the disk device to avoid risks associated with the disk replacement operation.
In the Isolate Damaged Disk step, click OK. Wait for isolation to complete. If the page shows an Instance Restart step, restart the instance.

Note
If the next step does not appear after a few minutes, refresh the page.

API operation

Query system events.

Call DescribeInstanceHistoryEvents to query system events in the Inquiring (Inquiring) state in the specified region. Record the returned EventId (EventId) and event name (Name).
Authorize and execute the isolation of damaged disks.

Call AcceptInquiredSystemEvent to authorize disk isolation.
- When the event name is SystemMaintenance.IsolateErrorDisk:
  - If only RequestId is returned, no restart is required. Wait for Alibaba Cloud to replace the disk.
  - If code:SwitchToOffline.OnlineIsolateFail is returned, restart the instance.
- When the event name is SystemMaintenance.RebootAndIsolateErrorDisk: After calling AcceptInquiredSystemEvent, restart the instance.

Note

After restart, the isolated damaged local disk becomes a 1 MiB dummy disk. Continue to isolate read/write operations on the damaged disk at the application layer and keep the nofail and barrier settings in /etc/fstab.

Replace the disk

No action is required. Wait for Alibaba Cloud to replace the damaged local disk. Replacement is expected within five business days. Check progress on the Local Disk Events page.

Restore the disk

ECS console

When Disk Repair Progress changes to To Be Restored, indicating that the event is in the Restore stage, click Repair in the Actions column, then click Restore in the dialog box.
If the page shows a Restart step, restart the instance.
When Disk Repair Progress changes to Handled, the event is fully processed.

API

Query system events.

Call DescribeInstanceHistoryEvents to query system events. After disk replacement, Alibaba Cloud publishes a SystemMaintenance.ReInitErrorDisk or SystemMaintenance.RebootAndReInitErrorDisk event.
Authorize disk restoration.

Call AcceptInquiredSystemEvent to authorize disk restoration.
- When the event name is SystemMaintenance.ReinitErrorDisk:
  - If only RequestId is returned, no restart is required.
  - If code=SwitchToOffline.OnlineReInitFail is returned, restart the instance.
- When the event name is SystemMaintenance.RebootAndReinitErrorDisk: After calling AcceptInquiredSystemEvent, restart the instance.

Next steps

After repair, check the ECS instance and local disk status. The replaced local disk must be initialized before use. See Initialize Windows data disk or Initialize Linux data disk.