Shahar Havivi
2015-Jan-20 14:25 UTC
Re: [Libguestfs] [PATCH] v2v: -o vdsm should assume data domain at -os path
On 20.01.15 12:59, Richard W.M. Jones wrote:> On Tue, Jan 20, 2015 at 02:53:43PM +0200, Shahar Havivi wrote: > > > >From 87f448fc8b3243ad5d01acb65efb0c710d811a93 Mon Sep 17 00:00:00 2001 > > From: Shahar Havivi <shaharh@redhat.com> > > Date: Tue, 20 Jan 2015 14:41:23 +0200 > > Subject: [PATCH] v2v: -o vdsm should assume data domain at -os path > > > > Unlike -o rhev which have only one data domin, -o rhev can and usually > > ^^ vdsm? > > > does have multiple data domain. > > The path to vdsm is pre mounted so no need to assume nfs path with -os > > Example: > > -o vdsm -os /rhev/data-center/<data-center-uuid>/<data-doamin-uuid> > > s/doamin/domain/ > > > > > Bug-Url: https://bugzilla.redhat.com/1176591 > > Signed-off-by: Shahar Havivi <shaharh@redhat.com> > > --- > > v2v/output_vdsm.ml | 10 +++++----- > > v2v/virt-v2v.pod | 3 ++- > > 2 files changed, 7 insertions(+), 6 deletions(-) > > > > diff --git a/v2v/output_vdsm.ml b/v2v/output_vdsm.ml > > index 492f586..17f10a5 100644 > > --- a/v2v/output_vdsm.ml > > +++ b/v2v/output_vdsm.ml > > @@ -63,9 +63,8 @@ object > > * name of the target files that eventually get written by the main > > * code. > > * > > - * 'os' is the output storage (-os nfs:/export). 'source' contains a > > - * few useful fields such as the guest name. 'targets' describes the > > - * destination files. We modify and return this list. > > + * 'os' is the output storage domain (-os /rhev/data/<data center>/<data domain>) > > + * this is already mounted path. > > * > > * Note it's good to fail here (early) if there are any problems, since > > * the next time we are called (in {!create_metadata}) we have already > > @@ -78,8 +77,9 @@ object > > error (f_"the number of '--vdsm-image-uuid' and '--vdsm-vol-uuid' parameters passed on the command line has to match the number of guest disk images (for this guest: %d)") > > (List.length targets); > > > > - let mp, uuid > > - Output_rhev.mount_and_check_storage_domain verbose (s_"Data Domain") os in > > + let pos = (String.length os) - 36 in > > + let mp = String.sub os 0 (pos-1) in > > + let uuid = String.sub os pos 36 in > > What is the significance of 36 here?the length of uuid.> > > dd_mp <- mp; > > dd_uuid <- uuid; > > if verbose then > > diff --git a/v2v/virt-v2v.pod b/v2v/virt-v2v.pod > > index a48bf59..cfd55fc 100644 > > --- a/v2v/virt-v2v.pod > > +++ b/v2v/virt-v2v.pod > > @@ -456,7 +456,8 @@ See L</OUTPUT TO RHEV> below. > > > > Set the output method to I<vdsm>. > > > > -This mode is similar to I<-o rhev> but is only used by RHEV VDSM > > +This mode is similar to I<-o rhev> but assume data domain as well > > +as data center /rhev/data-center/<data-center-uuid>/<data-domain-uuid> > > when it runs virt-v2v under VDSM control. > > You can't use < and > in POD. > > Probably better to phrase this as: > > This mode is similar to I<-o rhev>, but the full path to the > data domain must be given: > C</rhev/data-center/E<lt>data-center-uuidE<gt>/E<lt>data-domain-uuidE<gt>>. > This mode is only used when virt-v2v runs under VDSM control. > > Rich. > > -- > Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones > Read my programming and virtualization blog: http://rwmj.wordpress.com > virt-builder quickly builds VMs from scratch > http://libguestfs.org/virt-builder.1.htmlSure
Shahar Havivi
2015-Jan-20 14:39 UTC
Re: [Libguestfs] [PATCH] v2v: -o vdsm should assume data domain at -os path
Richard W.M. Jones
2015-Jan-20 14:48 UTC
Re: [Libguestfs] [PATCH] v2v: -o vdsm should assume data domain at -os path
On Tue, Jan 20, 2015 at 04:25:20PM +0200, Shahar Havivi wrote:> > > + let pos = (String.length os) - 36 in > > > + let mp = String.sub os 0 (pos-1) in > > > + let uuid = String.sub os pos 36 in > > > > What is the significance of 36 here? > the length of uuid.That code above will throw an exception if you pass a too-short -os parameter. How about this, which should be a little bit easier to understand: let mp, uuid let fields = string_nsplit "/" os in (* ... "data-center" "UUID" *) let fields = List.rev fields in (* "UUID" "data-center" ... *) match fields with | "" :: uuid :: rest (* handles trailing "/" case *) | uuid :: rest when String.length uuid = 36 -> let mp = String.concat "/" (List.rev rest) in mp, uuid | _ -> error (f_"vdsm: invalid -os parameter does not contain a valid UUID: %s") os Or perhaps something with regexps ... Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org
Shahar Havivi
2015-Jan-20 15:02 UTC
Re: [Libguestfs] [PATCH] v2v: -o vdsm should assume data domain at -os path
On 20.01.15 14:48, Richard W.M. Jones wrote:> On Tue, Jan 20, 2015 at 04:25:20PM +0200, Shahar Havivi wrote: > > > > + let pos = (String.length os) - 36 in > > > > + let mp = String.sub os 0 (pos-1) in > > > > + let uuid = String.sub os pos 36 in > > > > > > What is the significance of 36 here? > > the length of uuid. > > That code above will throw an exception if you pass a too-short -os > parameter. > > How about this, which should be a little bit easier to understand: > > let mp, uuid > let fields = string_nsplit "/" os in (* ... "data-center" "UUID" *) > let fields = List.rev fields in (* "UUID" "data-center" ... *) > match fields with > | "" :: uuid :: rest (* handles trailing "/" case *) > | uuid :: rest > when String.length uuid = 36 -> > let mp = String.concat "/" (List.rev rest) in > mp, uuid > | _ -> > error (f_"vdsm: invalid -os parameter does not contain a valid UUID: %s") > os >Looks better, I will give it a try> Or perhaps something with regexps ... > > Rich. > > -- > Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones > Read my programming and virtualization blog: http://rwmj.wordpress.com > libguestfs lets you edit virtual machines. Supports shell scripting, > bindings from many languages. http://libguestfs.org
Possibly Parallel Threads
- Re: [PATCH] v2v: -o vdsm should assume data domain at -os path
- Re: [PATCH] v2v: -o vdsm should assume data domain at -os path
- Re: [PATCH] v2v: -o vdsm should assume data domain at -os path
- Re: [PATCH] v2v: -o vdsm should assume data domain at -os path
- Re: [PATCH] v2v: -o vdsm should assume data domain at -os path