- Bugfixes only, no new features
- Allow online change of al-stripes and al-stripe-size
- Support for discard (ATA: trim, SCSI: unmap).
Drbd-9.0 is already available as prerelease, but not yet stable. Therefore drbd-9.0's features are still listed here on the ROADMAP.
- More than two nodes in one level (i.e. without device stacking)
- Automatic promote
- Striped activity log and non blocking AL updates
- Redundant connections between two nodes
- Support for request-relaying (remove the requirement for a full mesh)
- Support clusters where a partition has access to a shared disk
- Support for more than two primary nodes
- UDP multicast transport (instead of TCP connection mesh)
- Create new bitmap-slots online (online grow of backing storage is a pre condition)
Ideas, not yet planed for a specific release
- Never report false-positives with the data-integrity-alg checks.
- Use the crypto API in a way that we can utilize multiple CPU cores to do checksums
- Support DIF and DIX.
- Optionally deal with read errors in a way that the BBR (bad block relocation) of a drive can kick in
- move the cpu-mask from the syncer section to the host section, since it might be necessary to have that different on the systems.
- Offer an improved resync mode, that deals gracefully with thin provisioned storage. Currently a checksum based resync already has this property. The improvement will be a special mode, that just does not rewrite blocks containing only zeros.