Tools » teuthology

I've got a commit pushed that makes rpm based systems use qemu-kvm:

https://github.com/ceph/ceph-qa-suite/commit/c34dfd6d31e97bc50fb68286924c6f228f1e7405

I scheduled a test run using this branch, but the job still failed. I'm assuming there is some difference in the flags between qemu-kvm and qemu-system-x86_64. So far, I've noticed that the -drive flag looks to be different. Specifically, qemu-kvm doesn't have the 'format' option for -drive.

http://pulpito.front.sepia.ceph.com/ubuntu-2014-12-22_09:20:53-rbd-next-distro-basic-multi/672411/

Actually, after more testing it does look like the command flags are the same between the two version of qemu. The command is still failing though.

This is the output I get when running this command directly on a node.

[ubuntu@magna107 ~]$ adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage daemon-helper term /usr/libexec/qemu-kvm -enable-kvm -nographic -m 4096 -drive file=/home/ubuntu/cephtest/qemu/base.client.0.qcow2,format=qcow2,if=virtio -cdrom /home/ubuntu/cephtest/qemu/client.0.iso -fsdev local,id=log,path=/home/ubuntu/cephtest/archive/qemu/client0,secu
rity_model=none -device virtio-9p-pci,fsdev=log,mount_tag=test_log -drive file=rbd:rbd/client.0.0:id=0,format=raw,if=virtio,cache=none
qemu-kvm: -drive file=rbd:rbd/client.0.0:id=0,format=raw,if=virtio,cache=none: could not open disk image rbd:rbd/client.0.0:id=0: Unknown protocol
daemon-helper: command failed with exit status 1

I got this output by using interactive-on-error: true, modifying the task to abort and print the command before cleanup happens and then manually running the command on the node.

When I run `qemu-img -h` on magna107, it does not look like we're using a version of qemu-kvm with rbd support.

[ubuntu@magna107 ~]$ qemu-img -h
qemu-img version 1.5.3, Copyright (c) 2004-2008 Fabrice Bellard
usage: qemu-img command [command options]
QEMU disk image utility

Command syntax:
  check [-q] [-f fmt] [--output=ofmt]  [-r [leaks | all]] filename
  create [-q] [-f fmt] [-o options] filename [size]
  commit [-q] [-f fmt] [-t cache] filename
  compare [-f fmt] [-F fmt] [-p] [-q] [-s] filename1 filename2
  convert [-c] [-p] [-q] [-n] [-f fmt] [-t cache] [-O output_fmt] [-o options] [-s snapshot_name] [-S sparse_size] filename [filename2 [...]] output_filename
  info [-f fmt] [--output=ofmt] [--backing-chain] filename
  map [-f fmt] [--output=ofmt] filename
  snapshot [-q] [-l | -a snapshot | -c snapshot | -d snapshot] filename
  rebase [-q] [-f fmt] [-t cache] [-p] [-u] -b backing_file [-F backing_fmt] filename
  resize [-q] filename [+ | -]size
  amend [-q] [-f fmt] -o options filename

Command parameters:
  'filename' is a disk image filename
  'fmt' is the disk image format. It is guessed automatically in most cases
  'cache' is the cache mode used to write the output disk image, the valid
    options are: 'none', 'writeback' (default, except for convert), 'writethrough',
    'directsync' and 'unsafe' (default for convert)
  'size' is the disk image size in bytes. Optional suffixes
    'k' or 'K' (kilobyte, 1024), 'M' (megabyte, 1024k), 'G' (gigabyte, 1024M)
    and T (terabyte, 1024G) are supported. 'b' is ignored.
  'output_filename' is the destination disk image filename
  'output_fmt' is the destination format
  'options' is a comma separated list of format specific options in a
    name=value format. Use -o ? for an overview of the options supported by the
    used format
  '-c' indicates that target image must be compressed (qcow format only)
  '-u' enables unsafe rebasing. It is assumed that old and new backing file
       match exactly. The image doesn't need a working backing file before
       rebasing in this case (useful for renaming the backing file)
  '-h' with or without a command shows this help and lists the supported formats
  '-p' show progress of command (only certain commands)
  '-q' use Quiet mode - do not print any output (except errors)
  '-S' indicates the consecutive number of bytes (defaults to 4k) that must
       contain only zeros for qemu-img to create a sparse image during
       conversion. If the number of bytes is 0, the source will not be scanned for
       unallocated or zero sectors, and the destination image will always be
       fully allocated
  '--output' takes the format in which the output must be done (human or json)
  '-n' skips the target volume creation (useful if the volume is created
       prior to running qemu-img)

Parameters to check subcommand:
  '-r' tries to repair any inconsistencies that are found during the check.
       '-r leaks' repairs only cluster leaks, whereas '-r all' fixes all
       kinds of errors, with a higher risk of choosing the wrong fix or
       hiding corruption that has already occurred.

Parameters to snapshot subcommand:
  'snapshot' is the name of the snapshot to create, apply or delete
  '-a' applies a snapshot (revert disk to saved state)
  '-c' creates a snapshot
  '-d' deletes a snapshot
  '-l' lists all snapshots in the given image

Parameters to compare subcommand:
  '-f' first image format
  '-F' second image format
  '-s' run in Strict mode - fail on different image size or sector allocation

Supported formats: vvfat vpc vmdk vhdx vdi sheepdog sheepdog sheepdog raw host_cdrom host_floppy host_device file qed qcow2 qcow parallels nbd nbd nbd iscsi gluster gluster gluster
gluster dmg cow cloop bochs blkverify blkdebug

I believe we need to build qemu-kvm with the `--enable-rbd` flag.

Output from qemu-kvm shows it doesn't support rbd as a drive format either.

[ubuntu@magna107 ~]$ /usr/libexec/qemu-kvm -drive format=?
Supported formats: vvfat vpc vmdk vhdx vdi sheepdog sheepdog sheepdog raw host_cdrom host_floppy host_device file qed qcow2 qcow parallels nbd nbd nbd iscsi gluster gluster gluster gluster dmg cow cloop bochs blkverify blkdebug

For posterity, here's a good reference: http://www.sebastien-han.fr/blog/2014/01/17/check-if-kvm-has-the-support-of-ceph/

Filed this ticket for the rbd support, http://tracker.ceph.com/issues/10480

https://github.com/ceph/ceph-qa-suite/pull/283