The storage service uses a fixed-size metadata ID space that can represent a maximum of 2 PiB of logical space. When a virtual volume is created, it immediately occupies the corresponding logical space regardless of whether data is actually written to it. Snapshots and fast clone volumes do not consume additional logical space until real data writes occur in the corresponding data regions. When the logical space of the cluster approaches exhaustion, new virtual volumes cannot be created and new data cannot be written.
The logical capacity used by the cluster has exceeded the threshold of { .threshold }.
When there is still a certain amount of remaining logical capacity before reaching 2 PiB (at the TiB level), there is no significant impact on system operation, except that new volumes cannot be created if they would exceed the remaining logical space.
When logical capacity approaches or reaches the 2 PiB limit, the system:
Cannot create new virtual volumes.
Experiences significantly increased I/O latency on running virtual volumes, or even I/O errors.
The used logical space in the cluster is approaching the system limit.
Delete some older snapshots or other lower-priority virtual volumes.
Migrate virtual machines or virtual volumes to another cluster.
Note:
After deleting snapshots or virtual volumes, logical space is typically not released until approximately 3 minutes later.
