[Engine-devel] Question about Disk Hotplug
--_000_EF26FC1776F7FF46BFC072393EFD13A29A4FC8SERV070corpeldora_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Hi everyone, We proposed some changes in how to query if an operating system supports ho= tplug, then Itamar suggested that disk hotplugging and NIC hotplugging coul= d be separated from each other. While it is pretty clear how to put info ab= out NIC hotplugging in the osinfo, here are a few issues when trying to int= roduce the same concept for disk hotplugging. Some disk interfaces do not support hotplugging at all (like IDE), some of = them support it in the x86_64, but not in the ppc64 architecture (like the = VirtIO blk) and others support it in both architectures (like VirtIO SCSI).= There is also a ppc64 specific controller called SPAPR VSCSI, which suppor= ts it too. So, how do you propose a syntax to cover all these cases? I thought about something along the lines of: os.{id}.devices.disk.hotpluggableInterfaces =3D VirtIO_SCSI, VirtIO Thanks, Leonardo Bianconi/Vitor de Lima/Gustavo Pedrosa --_000_EF26FC1776F7FF46BFC072393EFD13A29A4FC8SERV070corpeldora_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable <html xmlns:v=3D"urn:schemas-microsoft-com:vml" xmlns:o=3D"urn:schemas-micr= osoft-com:office:office" xmlns:w=3D"urn:schemas-microsoft-com:office:word" = xmlns:m=3D"http://schemas.microsoft.com/office/2004/12/omml" xmlns=3D"http:= //www.w3.org/TR/REC-html40"> <head> <meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Dus-ascii"=
<meta name=3D"Generator" content=3D"Microsoft Word 14 (filtered medium)"> <style><!-- /* Font Definitions */ @font-face {font-family:Calibri; panose-1:2 15 5 2 2 2 4 3 2 4;} /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {margin:0cm; margin-bottom:.0001pt; font-size:11.0pt; font-family:"Calibri","sans-serif"; mso-fareast-language:EN-US;} a:link, span.MsoHyperlink {mso-style-priority:99; color:blue; text-decoration:underline;} a:visited, span.MsoHyperlinkFollowed {mso-style-priority:99; color:purple; text-decoration:underline;} span.EmailStyle17 {mso-style-type:personal-compose; font-family:"Calibri","sans-serif"; color:windowtext;} .MsoChpDefault {mso-style-type:export-only; font-family:"Calibri","sans-serif"; mso-fareast-language:EN-US;} @page WordSection1 {size:612.0pt 792.0pt; margin:70.85pt 3.0cm 70.85pt 3.0cm;} div.WordSection1 {page:WordSection1;} --></style><!--[if gte mso 9]><xml> <o:shapedefaults v:ext=3D"edit" spidmax=3D"1026" /> </xml><![endif]--><!--[if gte mso 9]><xml> <o:shapelayout v:ext=3D"edit"> <o:idmap v:ext=3D"edit" data=3D"1" /> </o:shapelayout></xml><![endif]--> </head> <body lang=3D"PT-BR" link=3D"blue" vlink=3D"purple"> <div class=3D"WordSection1"> <p class=3D"MsoNormal"><span lang=3D"EN-US">Hi everyone,<o:p></o:p></span><= /p> <p class=3D"MsoNormal"><span lang=3D"EN-US"><o:p> </o:p></span></p> <p class=3D"MsoNormal"><span lang=3D"EN-US">We proposed some changes in how= to query if an operating system supports hotplug, then Itamar suggested th= at disk hotplugging and NIC hotplugging could be separated from each other.= While it is pretty clear how to put info about NIC hotplugging in the osinfo, here are a few issues when tryin= g to introduce the same concept for disk hotplugging.<o:p></o:p></span></p> <p class=3D"MsoNormal"><span lang=3D"EN-US"><o:p> </o:p></span></p> <p class=3D"MsoNormal"><span lang=3D"EN-US">Some disk interfaces do not sup= port hotplugging at all (like IDE), some of them support it in the x86_64, = but not in the ppc64 architecture (like the VirtIO blk) and others support = it in both architectures (like VirtIO SCSI). There is also a ppc64 specific controller called SPAPR VSCSI, which= supports it too.<o:p></o:p></span></p> <p class=3D"MsoNormal"><span lang=3D"EN-US"><o:p> </o:p></span></p> <p class=3D"MsoNormal"><span lang=3D"EN-US">So, how do you propose a syntax= to cover all these cases?<o:p></o:p></span></p> <p class=3D"MsoNormal"><span lang=3D"EN-US"><o:p> </o:p></span></p> <p class=3D"MsoNormal"><span lang=3D"EN-US">I thought about something along= the lines of:<o:p></o:p></span></p> <p class=3D"MsoNormal"><span lang=3D"EN-US">os.{id}.devices.disk.hotpluggab= leInterfaces =3D VirtIO_SCSI, VirtIO<o:p></o:p></span></p> <p class=3D"MsoNormal"><span lang=3D"EN-US"><o:p> </o:p></span></p> <p class=3D"MsoNormal">Thanks,<o:p></o:p></p> <p class=3D"MsoNormal">Leonardo Bianconi/Vitor de Lima/Gustavo Pedrosa<o:p>= </o:p></p> </div> </body> </html> --_000_EF26FC1776F7FF46BFC072393EFD13A29A4FC8SERV070corpeldora_--
On Tue 10 Sep 2013 09:20:32 PM IDT, Gustavo Frederico Temple Pedrosa wrote:
Hi everyone,
We proposed some changes in how to query if an operating system supports hotplug, then Itamar suggested that disk hotplugging and NIC hotplugging could be separated from each other. While it is pretty clear how to put info about NIC hotplugging in the osinfo, here are a few issues when trying to introduce the same concept for disk hotplugging.
Some disk interfaces do not support hotplugging at all (like IDE), some of them support it in the x86_64, but not in the ppc64 architecture (like the VirtIO blk) and others support it in both architectures (like VirtIO SCSI). There is also a ppc64 specific controller called SPAPR VSCSI, which supports it too.
So, how do you propose a syntax to cover all these cases?
I thought about something along the lines of:
os.{id}.devices.disk.hotpluggableInterfaces = VirtIO_SCSI, VirtIO
Thanks,
Leonardo Bianconi/Vitor de Lima/Gustavo Pedrosa
"storage" is a better term than "disk" so: os.{id}.devices.storage... maybe in the future we may expand the metadata model to hold devices and hook it to an OS device.{id}.... os.{id}.devices={deviceId} just a thought, its unnecessary for now.
On 09/11/2013 11:27 AM, Roy Golan wrote:
On Tue 10 Sep 2013 09:20:32 PM IDT, Gustavo Frederico Temple Pedrosa wrote:
Hi everyone,
We proposed some changes in how to query if an operating system supports hotplug, then Itamar suggested that disk hotplugging and NIC hotplugging could be separated from each other. While it is pretty clear how to put info about NIC hotplugging in the osinfo, here are a few issues when trying to introduce the same concept for disk hotplugging.
Some disk interfaces do not support hotplugging at all (like IDE), some of them support it in the x86_64, but not in the ppc64 architecture (like the VirtIO blk) and others support it in both architectures (like VirtIO SCSI). There is also a ppc64 specific controller called SPAPR VSCSI, which supports it too.
So, how do you propose a syntax to cover all these cases?
I thought about something along the lines of:
os.{id}.devices.disk.hotpluggableInterfaces = VirtIO_SCSI, VirtIO
Thanks,
Leonardo Bianconi/Vitor de Lima/Gustavo Pedrosa
"storage" is a better term than "disk" so:
why? this is about a vdisk, isn't it?
os.{id}.devices.storage...
maybe in the future we may expand the metadata model to hold devices and hook it to an OS
device.{id}.... os.{id}.devices={deviceId}
just a thought, its unnecessary for now.
_______________________________________________ Engine-devel mailing list Engine-devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/engine-devel
On Wed 11 Sep 2013 12:06:40 PM IDT, Itamar Heim wrote:
On 09/11/2013 11:27 AM, Roy Golan wrote:
On Tue 10 Sep 2013 09:20:32 PM IDT, Gustavo Frederico Temple Pedrosa wrote:
Hi everyone,
We proposed some changes in how to query if an operating system supports hotplug, then Itamar suggested that disk hotplugging and NIC hotplugging could be separated from each other. While it is pretty clear how to put info about NIC hotplugging in the osinfo, here are a few issues when trying to introduce the same concept for disk hotplugging.
Some disk interfaces do not support hotplugging at all (like IDE), some of them support it in the x86_64, but not in the ppc64 architecture (like the VirtIO blk) and others support it in both architectures (like VirtIO SCSI). There is also a ppc64 specific controller called SPAPR VSCSI, which supports it too.
So, how do you propose a syntax to cover all these cases?
I thought about something along the lines of:
os.{id}.devices.disk.hotpluggableInterfaces = VirtIO_SCSI, VirtIO
Thanks,
Leonardo Bianconi/Vitor de Lima/Gustavo Pedrosa
"storage" is a better term than "disk" so:
why? this is about a vdisk, isn't it? well a vnic and vdisk are the building blocks in the end. so either we use vnic vdisk etc or storage,network etc for all
os.{id}.devices.storage...
maybe in the future we may expand the metadata model to hold devices and hook it to an OS
device.{id}.... os.{id}.devices={deviceId}
just a thought, its unnecessary for now.
_______________________________________________ Engine-devel mailing list Engine-devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/engine-devel
On 09/11/2013 04:51 PM, Roy Golan wrote:
On Wed 11 Sep 2013 12:06:40 PM IDT, Itamar Heim wrote:
On 09/11/2013 11:27 AM, Roy Golan wrote:
On Tue 10 Sep 2013 09:20:32 PM IDT, Gustavo Frederico Temple Pedrosa wrote:
Hi everyone,
We proposed some changes in how to query if an operating system supports hotplug, then Itamar suggested that disk hotplugging and NIC hotplugging could be separated from each other. While it is pretty clear how to put info about NIC hotplugging in the osinfo, here are a few issues when trying to introduce the same concept for disk hotplugging.
Some disk interfaces do not support hotplugging at all (like IDE), some of them support it in the x86_64, but not in the ppc64 architecture (like the VirtIO blk) and others support it in both architectures (like VirtIO SCSI). There is also a ppc64 specific controller called SPAPR VSCSI, which supports it too.
So, how do you propose a syntax to cover all these cases?
I thought about something along the lines of:
os.{id}.devices.disk.hotpluggableInterfaces = VirtIO_SCSI, VirtIO
Thanks,
Leonardo Bianconi/Vitor de Lima/Gustavo Pedrosa
"storage" is a better term than "disk" so:
why? this is about a vdisk, isn't it? well a vnic and vdisk are the building blocks in the end. so either we use vnic vdisk etc or storage,network etc for all
we could have other "storage" things than disks?
os.{id}.devices.storage...
maybe in the future we may expand the metadata model to hold devices and hook it to an OS
device.{id}.... os.{id}.devices={deviceId}
just a thought, its unnecessary for now.
_______________________________________________ Engine-devel mailing list Engine-devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/engine-devel
participants (3)
-
Gustavo Frederico Temple Pedrosa -
Itamar Heim -
Roy Golan