Bring SCSI Support Into QEMU Block Layer
summary
Adds SCSI reservation support directly into QEMU’s block layer so guests access Ceph-like shared disks without iSCSI, implementing four components: device emulation, block-backend interface, block-driver interface and local driver hooks. Implements compare-and-write and a persistent reservation API, adds four new block-driver calls (including RBD hooks), and reuses existing block EIO paths. Compare-and-write patches are upstream; persistent reservation is pending. The patchset runs in a cloud deployment and functions normally.
AI-written summary · about
chapters (12)
- 0:00Intro
- 0:25Outline
- 1:00Background
- 1:40Architecture
- 2:15Problems
- 3:00Solution
- 4:35Implementation
- 5:10Scarcity Devices
- 6:20Pass Interface
- 7:00Block Driver Interface
- 7:35Demonstration
- 8:20Future Plan
related talks
30mQEMU-Storage-Daemon and libblkio: Exploring New Shores for the QE...
20mDeep Dive into Ceph Block Storage
34mIn-Guest Mechanism to Strengthen Guest Separation
44mVM block error injection, a novel approach for testing Linux storage
03mZoned MC Conference Kick Off
28mCXL Emulation in QEMU - Progress, status and most importantly what next?
18mDm-qcow2: qcow2 as Device-mapper Block Device
21mRISC V Hypervisor Support