similar to: [PATCH] v2v: adding --vdsm-ovf-output option

This option is needed by vdsm for writing the ovf to a specific directory. The default is current directory. Bug-Url: https://bugzilla.redhat.com/show_bug.cgi?id=1176598 Signed-off-by: Shahar Havivi <shaharh@redhat.com> --- v2v/cmdline.ml | 5 +++++ v2v/output_vdsm.ml | 16 +++++++--------- v2v/output_vdsm.mli | 1 + v2v/test-v2v-o-vdsm-options.sh |

This option is needed by vdsm for writing the ovf to a specific directory. The default is current directory. Bug-Url: https://bugzilla.redhat.com/show_bug.cgi?id=1176598 Signed-off-by: Shahar Havivi <shaharh@redhat.com> --- v2v/cmdline.ml | 5 +++++ v2v/output_vdsm.ml | 16 +++++++--------- v2v/output_vdsm.mli | 1 + v2v/test-v2v-o-vdsm-options.sh |

This option is needed by vdsm for writing the ovf to a specific directory. The default is current directory. Bug-Url: https://bugzilla.redhat.com/show_bug.cgi?id=1176598 Signed-off-by: Shahar Havivi <shaharh@redhat.com> --- v2v/cmdline.ml | 5 +++++ v2v/output_vdsm.ml | 16 +++++++--------- v2v/output_vdsm.mli | 1 + v2v/test-v2v-o-vdsm-options.sh |

This option is needed by vdsm for writing the ovf to a specific directory. The default is current directory. Bug-Url: https://bugzilla.redhat.com/show_bug.cgi?id=1176598 Signed-off-by: Shahar Havivi <shaharh@redhat.com> --- v2v/cmdline.ml | 5 +++++ v2v/output_vdsm.ml | 16 +++++++--------- v2v/output_vdsm.mli | 1 + v2v/test-v2v-o-vdsm-options.sh |

The main reason for creating different OVF is that it can be used to create VM by oVirt REST API. The RHV export domain flavor cannot be used that way. Tomáš Golembiovský (3): v2v: tests: check generated OVF v2v: ovf: Create OVF more aligned with the standard v2v: vdsm: add --vdsm-fixed-ovf option v2v/cmdline.ml | 5 ++ v2v/create_ovf.ml

The main reason for creating different OVF is that it can be used to create VM by oVirt REST API. The RHV export domain flavor cannot be used that way. v1 -> v2: - introduced flavour types instead of booleans - instead of referring to the new flavour as "standard OVF" or "fixed OVF" I refer to it as oVirt flavour. While it is more conforming than the one used in export

v3: Change the flag from --vdsm-compat-11 to --vdsm-compat=1.1 Also the --machine-readable output has changed. I have also added a test. Rich.

Add option for -o vdsm that enables output of the modified OVF. oVirt engine should already be able to consume the OVF, but let's not take any chances and enable it only by command line argument. It can be made default later when it receives proper testing. Signed-off-by: Tomáš Golembiovský <tgolembi@redhat.com> --- v2v/cmdline.ml | 5 +++++

This patch isn't quite complete (see ‘assert false’). Currently we have a bunch of ad hoc options like --vddk* and --vdsm* (and proposed to add --rhv*) to handle extra parameters for input and output modes/transports. This complicates the command line parsing and also the clarity of the command line (becauseit's not very obvious which options apply to which side of the conversion).

The first two patches implement the change discussed here: https://www.redhat.com/archives/libguestfs/2016-April/msg00178.html The third patch fixes the mapping of inspection data to OVF VmType which was inherited directly from old virt-v2v and had never been changed. It had a number of problems. The fourth patch is only slightly related to the others. It adds an extra slow test to ensure that

Support for RHEV with RHEL 6 nodes required us to output the old style qcow2 compat=0.10 images. Since RHEV 3.6 GA, RHEL 6 has not been supported as a RHEV node type. Since RHV 4.1, compat=1.1 is supported. Support for compat=1.1 is uncertain in RHV 4.0 even on RHEL 7 nodes. There are significant downsides to using qcow2 compat=0.10 instead of the modern default (compat=1.1). Therefore this

Support for RHEV with RHEL 6 nodes required us to output the old style qcow2 compat=0.10 images. Since RHEV 3.6 GA, RHEL 6 has not been supported as a RHEV node type. Since RHV 4.1, compat=1.1 is supported. (Support for compat=1.1 is uncertain in RHV 4.0 even on RHEL 7 nodes.) There are significant downsides to using qcow2 compat=0.10 instead of the modern default (compat=1.1). Therefore this

v6 was here: https://www.redhat.com/archives/libguestfs/2018-March/msg00126.html This makes a number of significant changes: - Input and output options now use a uniform set of -io and -oo parameters. - For -o rhv-upload, we use ‘-oo rhv-cafile=/tmp/ca.pem’ etc. The ‘--rhv*’ options have been dropped. - Rearranges the documentation. - As before includes (untested) support for zero,

Currently we have a bunch of ad hoc options like --vddk* and --vdsm* (and proposed to add --rhv*) to handle extra parameters for input and output modes/transports. This complicates the command line parsing and also the clarity of the command line (becauseit's not very obvious which options apply to which side of the conversion). Replace these with a general mechanism for handling input and

I still have not found time to fully review the series, but I found out that this patch breaks VDDK. See below. On Thu, 22 Mar 2018 15:24:23 +0000 "Richard W.M. Jones" <rjones@redhat.com> wrote: > Currently we have a bunch of ad hoc options like --vddk* and --vdsm* > (and proposed to add --rhv*) to handle extra parameters for input and > output modes/transports. This

When erroring out about duplicated parameters, say "more than once" instead of "twice", since there can be more than two repeated parameters. Thanks to: Xiaodai Wang --- generator/fish.ml | 2 +- generator/perl.ml | 2 +- resize/resize.ml | 4 ++-- tools/virt-tar | 4 ++-- v2v/input_libvirt_vddk.ml | 2 +- v2v/output_rhv_upload.ml | 4 ++--

For historical reasons the OVF used in RHV export domain contains some deviations from the OVF standard. The format used in -o rhv has to remain fixed but for -o vdsm and we could produce much nicer OVF. This patch serves as a preparatory step to this. The main reason for creating different OVF is that it can be used to create VM by oVirt REST API. The RHV export domain flavor cannot be used that

When converting and throwing away the output (-o null), the input format should not really matter. However it does currently matter for a couple of reasons: (1) If the input is not raw or qcow2, then you'll get the error below because virt-v2v wants to write the same output format as input format but only allows raw or qcow2 as output formats: virt-v2v: error: output format should be

Don't pass these flags to dozens of functions. --- builder/builder.ml | 47 +++++++++-------- builder/cache.ml | 4 +- builder/cache.mli | 2 +- builder/cmdline.ml | 13 ++--- builder/downloader.ml | 14 +++-- builder/downloader.mli

.. and a lot of refactoring. https://bugzilla.redhat.com/show_bug.cgi?id=1167623 Rich.