[Bug] Major compaction is not recovered automatically after the disk is writable again

### Search before asking

- [X] I searched in the [issues](https://github.com/apache/pulsar/issues) and found nothing similar.

### Read release policy

- [X] I understand that unsupported versions don't get bug fixes. I will attempt to reproduce the issue on a supported version of Pulsar client and Pulsar broker.

### Version

v2.11.2

### Minimal reproduce step

1. Major compaction task is suspended because the disk is almost full.

2. Minor compaction task is also suspended because the disk is full-filled.

3. Disk becomes writable again.

**Problem1**: Only `fulfilledDirs` will be recovered. If the disks(dirs) are just almost full(not full-filled), there is no change their major compaction or minor compaction task can be recovered.

**Problem2**:
Because the disk's usage is still not under warn threshold (0.9) when they are writable, only minor compaction task can be recovered and then the disks(dirs) are removed from `fulfilledDirs`, So there is no change their major compaction task can be recovered because of `Problem1`.

4. Finally, the major compaction task can not be recovered even the disk is idle.

### What did you expect to see?

Major compaction is recoveryed automatically when the dist is idle again.

### What did you see instead?

Major compaction cannot be recoveryed automatically when the dist is idle again.

### Anything else?

_No response_

### Are you willing to submit a PR?

- [ ] I'm willing to submit a PR!

No source files, tests, or entry points are named. Start by reproducing the sequence in the issue on Pulsar v2.11.2: suspend major and minor compaction with a nearly full disk, restore writability, and observe recovery. Done means major compaction resumes automatically once the disk is idle again.

Written by the indexing model from the issue text.