RHEL Engineering is moving the tracking of its product development work on RHEL 6 through RHEL 9 to Red Hat Jira (issues.redhat.com). If you're a Red Hat customer, please continue to file support cases via the Red Hat customer portal. If you're not, please head to the "RHEL project" in Red Hat Jira and file new tickets here. Individual Bugzilla bugs in the statuses "NEW", "ASSIGNED", and "POST" are being migrated throughout September 2023. Bugs of Red Hat partners with an assigned Engineering Partner Manager (EPM) are migrated in late September as per pre-agreed dates. Bugs against components "kernel", "kernel-rt", and "kpatch" are only migrated if still in "NEW" or "ASSIGNED". If you cannot log in to RH Jira, please consult article #7032570. That failing, please send an e-mail to the RH Jira admins at rh-issues@redhat.com to troubleshoot your issue as a user management inquiry. The email creates a ServiceNow ticket with Red Hat. Individual Bugzilla bugs that are migrated will be moved to status "CLOSED", resolution "MIGRATED", and set with "MigratedToJIRA" in "Keywords". The link to the successor Jira issue will be found under "Links", have a little "two-footprint" icon next to it, and direct you to the "RHEL project" in Red Hat Jira (issue links are of type "https://issues.redhat.com/browse/RHEL-XXXX", where "X" is a digit). This same link will be available in a blue banner at the top of the page informing you that that bug has been migrated.
Bug 1200049 - VirtIO-SCSI device order wrong in VM
Summary: VirtIO-SCSI device order wrong in VM
Keywords:
Status: CLOSED WONTFIX
Alias: None
Product: Red Hat Enterprise Linux 6
Classification: Red Hat
Component: qemu-kvm
Version: 6.6
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: rc
: ---
Assignee: Fam Zheng
QA Contact: Virtualization Bugs
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2015-03-09 15:32 UTC by Evgheni Dereveanchin
Modified: 2019-04-16 14:41 UTC (History)
10 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2015-03-11 03:28:20 UTC
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)

Description Evgheni Dereveanchin 2015-03-09 15:32:38 UTC
Description of problem:
When using VirtIO-SCSI and multiple disks, the disk order visible in the VM is random, causing issues with VM provisioning (disk names unpredictable) and even unbootable VM (disk added becomes first in VM and renders it unbootable)

Version-Release number of selected component (if applicable):
qemu-img-rhev-0.12.1.2-2.446.el6.x86_64

RHEL 6.6 in VM

How reproducible:
Always

Steps to Reproduce:
1. Create a VM in RHEV with 4 disks:
Disk1 10GB
Disk2 20GB
Disk3 40GB
Disk4 30GB

2) use VirtIO bus, start VM and check disk proper names:
Disk1 /dev/vda
Disk2 /dev/vdb
Disk3 /dev/vdc
Disk4 /dev/vdd

3) attach all disks to the Virtio-SCSI controller
Actual results:
Inconsistent device naming:
Disk1 /dev/sdd
Disk2 /dev/sda
Disk3 /dev/sdb
Disk4 /dev/sdc
(may be in different order on a different VM)

Expected results:
Consistent device naming:
Disk1 /dev/sda
Disk2 /dev/sdb
Disk3 /dev/sdc
Disk4 /dev/sdd

Additional info:

here's what dmesg may look like in a VM:
sd 2:0:0:0: [sda] 20971520 512-byte logical blocks: (10.7 GB/10.0 GiB)
sd 2:0:0:2: [sdb] 62914560 512-byte logical blocks: (32.2 GB/30.0 GiB)
sd 2:0:0:3: [sdc] 83886080 512-byte logical blocks: (42.9 GB/40.0 GiB)
sd 2:0:0:1: [sdd] 41943040 512-byte logical blocks: (21.4 GB/20.0 GiB)

Comment 3 Fam Zheng 2015-03-11 03:28:20 UTC
A sane guest should not rely on the automatically generated /dev/sdX names to address a partition. Even if the order in a single bus is consistent, attaching a different bus (for example, ide/ahci) could as easily break the names.

Alternatively, there are /dev/disk/by-* symbol links that are more robust to system configuration changes. If you really need persistent /dev/sdX names, use udev persistent device name rules.

Fam

Comment 4 Fam Zheng 2015-03-11 03:32:26 UTC
By default, RHEL 6 installation (and almost every Linux distribution) uses UUID to find boot and root filesystem devices, so it will work still after you plug new disks that changes /dev/sdX order.

Comment 5 Guy Baconniere 2015-05-12 12:55:44 UTC
I had the same issue after switching from virtio-blk to virtio-scsi
udev does not cache sda pci map even if you add serial to the disk.

As workaround, I changed the target (on top of the unit) and 
it is working.

BEFORE

      <address type='drive' controller='0' bus='0' target='0' unit='0'/>
      <address type='drive' controller='0' bus='0' target='0' unit='1'/>
      <address type='drive' controller='0' bus='0' target='0' unit='2'/>
      <address type='drive' controller='0' bus='0' target='0' unit='3'/>

AFTER (workaround)

      <address type='drive' controller='0' bus='0' target='0' unit='0'/>
      <address type='drive' controller='0' bus='0' target='1' unit='1'/>
      <address type='drive' controller='0' bus='0' target='2' unit='2'/>
      <address type='drive' controller='0' bus='0' target='3' unit='3'/>

    <disk type='block' device='disk'>
      <driver name='qemu' type='raw' cache='writeback' discard='unmap'/>
      <source dev='/var/lib/libvirt/images/boot.raw'/>
      <target dev='sda' bus='scsi'/>
      <serial>42-001</serial>
      <boot order='1'/>
      <alias name='scsi0-0-0-0'/>
      <address type='drive' controller='0' bus='0' target='0' unit='0'/>
    </disk>
    <disk type='block' device='disk'>
      <driver name='qemu' type='raw' cache='writeback' discard='unmap'/>
      <source dev='/var/lib/libvirt/images/root.raw'/>
      <target dev='sdb' bus='scsi'/>
      <serial>42-002</serial>
      <alias name='scsi0-0-1-1'/>
      <address type='drive' controller='0' bus='0' target='1' unit='1'/>
    </disk>
    ...
    <controller type='scsi' index='0' model='virtio-scsi'>
      <alias name='scsi0'/>
      <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/>
    </controller>

The interessting part is also that the ID_SCSI does not change?!

for dev in sd{a..d}; do echo -e "\n[$dev]"; udevadm info --query=all --name=/dev/$dev | grep -E '(ID_SCSI|ID_PATH)'; done

[sda]
E: ID_PATH=pci-0000:00:05.0-virtio-pci-virtio2-scsi-0:0:0:0
E: ID_PATH_TAG=pci-0000_00_05_0-virtio-pci-virtio2-scsi-0_0_0_0
E: ID_SCSI=1
E: ID_SCSI_SERIAL=42-001

[sdb]
E: ID_PATH=pci-0000:00:05.0-virtio-pci-virtio2-scsi-0:0:1:1
E: ID_PATH_TAG=pci-0000_00_05_0-virtio-pci-virtio2-scsi-0_0_1_1
E: ID_SCSI=1
E: ID_SCSI_SERIAL=42-002

[sdc]
E: ID_PATH=pci-0000:00:05.0-virtio-pci-virtio2-scsi-0:0:2:2
E: ID_PATH_TAG=pci-0000_00_05_0-virtio-pci-virtio2-scsi-0_0_2_2
E: ID_SCSI=1
E: ID_SCSI_SERIAL=42-003

[sdd]
E: ID_PATH=pci-0000:00:05.0-virtio-pci-virtio2-scsi-0:0:3:3
E: ID_PATH_TAG=pci-0000_00_05_0-virtio-pci-virtio2-scsi-0_0_3_3
E: ID_SCSI=1
E: ID_SCSI_SERIAL=42-004

Comment 6 jmkim 2015-06-07 16:17:39 UTC
How do the possible need to modify a change?

Virtio-scsi must either modify the config volume may seem normal?


Note You need to log in before you can comment on or make changes to this bug.