Re: KVM question.

[Timothy Sipples]

Of course z/VM can run any/all IBM Z operating systems, including both 
z/VSE and Linux. And it can do so within even a single z/VM LPAR. There 
are some significant resource and operational efficiencies in that sort of 
configuration.

In 2017 IBM announced general availability of sub-capacity licensing for 
z/VM, so you can now license z/VM one engine at a time. Previously you had 
to license z/VM for all the IFLs, all the CPs, or both, per machine. Let's 
suppose for example you have 1 IFL and 3 CPs -- machine capacity model 
3907-C03 with 1 IFL as an example. You could configure a z/VM LPAR that 
spans the IFL and one CP (shared or dedicated), and that would require 
only 2 engines of z/VM licensing (down from the 4 previously required in 
this scenario). z/VSE could then run both in its own LPAR(s) and within 
the z/VM LPAR. Many variations are possible, of course, but that's one 
sample variation.

Either way (or both), I very much like the idea of using a second level 
hypervisor to run Linux, and to do so right at the beginning. Then you 
really don't have to give much thought to adding more Linux instances, 
even if the "new" Linux instances are for release upgrade reasons. It's 
not hard to do. In fact, in some ways it's easier to start off with a 
second level hypervisor.

- - - - - - - - - -
Timothy Sipples
I.T. Architect Executive
Digital Asset & Other Industry Solutions
IBM Z & LinuxONE
- - - - - - - - - -
E-Mail: sipp...@sg.ibm.com

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM question.

[Christian Borntraeger]

On 30.06.20 22:19, Tom Huegel wrote:
> This has probably been discussed quite a bit but I wanted to ask for some
> real user feedback.
>
> I have a customer that is interested a POC of z/VSE zLINUX VTAPE.
> Unfortunately they do now have z/VM so the zLINUX would have to be in an
> LPAR. So the question is 'do I gain/lose anything using KVM as the
> hypervisor in the zLINUX LPAR?'.

As the zLINUX VTAPE application is just a Java applicaition that needs network
and disk access I do not see a reason why it could not run in a KVM guest. I 
guess
you have to ask yourself the question, will there be additional things that you
might want to run on zLinux to support VSE. I think you got plenty of ideas
in the recent mail threads. Then of course a hypervisor is going to help you.
One nice thing with KVM is that you can actually add that later. So having one
workload in an LPAR and then addding a KVM guest in the same LPAR to host 
another
one would just work. Not that this is something that I would suggest.
But the point is: no matter if you run Linux in LPAR or Linux under KVM. You
always start with a Linux in LPAR and can then move to KVM or not.

With some preplanning for the future you probably want to put the first workload
also in a KVM guest to have equivalent management for all workloads (e.g. use
CPU shares or capping).
Regarding the downside, of course every hypervisor will add a small overhead. It
starts with the memory footprint of an additional Linux kernel and it ends with
the cost of virtualizing I/O. Does this matter for the vtape use case? Probably
not.
As an alternative that depends on your requirements regarding isolation you 
could
put the workloads in containers. But this is only "nice" when the workload is
already containerized.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM question.

[Jim Elliott]

Tom:

I believe you meant they do "not" have z/VM? If you are just looking at
this to run VTAPE you don't need a hypervisor. Run the Linux image in an
LPAR. My interpretation is that they don't have an IFL either so "native"
Linux would be easier to start with. If they decide to buy an IFL in the
future, IBM offers z/VM at a large discount when acquired with the IFL.
There are several attractive offerings to help z/VSE customers get started
with Linux on Z.

Jim Elliott
Senior IT Consultant - GlassHouse Systems Inc.


On Tue, Jun 30, 2020 at 4:21 PM Tom Huegel  wrote:

> This has probably been discussed quite a bit but I wanted to ask for some
> real user feedback.
>
> I have a customer that is interested a POC of z/VSE zLINUX VTAPE.
> Unfortunately they do now have z/VM so the zLINUX would have to be in an
> LPAR. So the question is 'do I gain/lose anything using KVM as the
> hypervisor in the zLINUX LPAR?'.
>
> Thanks
> Tom
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390
>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM question.

[Scott Rohling]

Others might be more specific about z/VM advantages over KVM ...  but you
certainly lose your knowledge of z/VM to manage and control guests and need
to use a KVM  based model for that.. whether that is an issue or not is
about your comfort level with both solutions on z.  (and probably others
you work with?).

Having said that - I've worked with IBM customers in many POCs where the
licensing for our solution was supplied on a trial basis - just for that
purpose -- so I'd think the z/VM licensing could be managed this way..

Sticking with a horse you know makes things quicker than learning to ride a
new one... and in my experience - expedience always seems to be a part of
any POC...   :)   And -- any solution you come up with can always get
shifted to an LPAR (or LPARs) fairly easily with the correct IOCONFIG for
the devices...  Use dedicated volumes under z/VM if you know you'll be
moving off it...  just in case you need to boot Linux native.   (or 0 to
END minidisks)

Scott Rohling

Best of luck - Scott

On Tue, Jun 30, 2020 at 1:20 PM Tom Huegel  wrote:

> This has probably been discussed quite a bit but I wanted to ask for some
> real user feedback.
>
> I have a customer that is interested a POC of z/VSE zLINUX VTAPE.
> Unfortunately they do now have z/VM so the zLINUX would have to be in an
> LPAR. So the question is 'do I gain/lose anything using KVM as the
> hypervisor in the zLINUX LPAR?'.
>
> Thanks
> Tom
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390
>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM only supports 4guests

[Mike Friesenegger]

Hello,

I just want to point out that a SUSE Linux Enterprise Server for IBM Z and 
LinuxONE subscription does NOT limit the number of KVM VMs.

Regards,

Mike
--
Mike Friesenegger
Solution Architect
IBM Global Alliance
SUSE
mi...@suse.com
(P)+1 303.249.0817

[https://www.suse.com/common/img/logo.png]

On 10/22/19 7:29 AM, Robert J Brenneman wrote:

Barton - you are correct.

RH only supports 'unlimited' numbers of virtual machines on their dedicated
hypervisor offering which is RHV ( used to be RHEV ).  RHV is not yet
available on s390x. I am not a part of the team that can say whether it
will become available for s390x, either, so your guess is as good as mine
about that.


On Tue, Oct 22, 2019 at 8:53 AM Bfishing 
 wrote:



I wonder if that winds up depending on where you or they run it?
- I stated it that way knowing that IBM is pushing KVM on zCloud over just
offering z/VM as a Service (zaas) for running Linux guest like they should
be!

On Mon, Oct 21, 2019 at 9:39 PM Barton 

wrote:



Right.  That’s what I’m hearing , that ‘unlimited ‘ is not available on Z

Barton




On Oct 22, 2019, at 12:27 AM, Jim Elliott 



wrote:



Sam:

The Red Hat site above does not make sense to me as there is only ONE
version of RHEL available for IBM Z. That is, Red Hat Enterprise Linux


with


Unlimited Guests and Red Hat Enterprise Virtualization are NOT


available


for IBM Z.

Jim Elliott
Senior Consultant - GlassHouse Systems Inc.





On Mon, Oct 21, 2019 at 6:07 PM Cohen, Sam 
 wrote:

Barton,

It depends on the licensed version of Redhat that you have.  From the
footnote:

This guest limit does not apply to Red Hat Enterprise Linux with


Unlimited


Guests. There is no guest limit for Red Hat Enterprise Virtualization,


Red


Hat Enterprise Linux OpenStack Platform, and the Red Hat Enterprise


Linux


with Smart Virtualization bundle.

Thanks,

Sam
(217) 862-9227 (office)
(602) 327-2134 (cell)

-Original Message-
From: Linux on 390 Port 
 On Behalf Of Barton
Sent: Monday, October 21, 2019 2:54 PM
To: LINUX-390@VM.MARIST.EDU
Subject: KVM only supports 4guests

I’m hearing at gse that red hat on z only supports 4 guests on kvm on


z?


Here’s a  link, am i missIng something?






https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Faccess.redhat.com%2Farticles%2Frhel-kvm-limitsdata=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=g%2FPSoSxYqh2SgqhLX5V5KrvdkgFPSuWcvqrkupgaKj0%3Dreserved=0



Barton


--
For LINUX-390 subscribe / signoff / archive access instructions, send
email to lists...@vm.marist.edu with the 
message: INFO LINUX-390 or


visit









https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww2.marist.edu%2Fhtbin%2Fwlvindex%3FLINUX-390data=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=imwx9PkkgF74LtoqU5u9d2mv2nZ0HgHu5OXuUlEwqVo%3Dreserved=0



--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the 
message: INFO LINUX-390


or


visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390




--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the 
message: INFO LINUX-390


or visit


http://www2.marist.edu/htbin/wlvindex?LINUX-390



--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the 
message: INFO LINUX-390 or
visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390





--



<º>`·.¸¸´¯`·.¸.·´¯`·...¸>(((º>


.·´¯`·.><º>`·.¸¸.·´¯`·.¸.·´¯`·...¸><º>

<>< Go fishing ><>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the 
message: INFO LINUX-390 or
visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390









--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM only supports 4guests

[Robert J Brenneman]

Barton - you are correct.

RH only supports 'unlimited' numbers of virtual machines on their dedicated
hypervisor offering which is RHV ( used to be RHEV ).  RHV is not yet
available on s390x. I am not a part of the team that can say whether it
will become available for s390x, either, so your guess is as good as mine
about that.


On Tue, Oct 22, 2019 at 8:53 AM Bfishing  wrote:

> I wonder if that winds up depending on where you or they run it?
> - I stated it that way knowing that IBM is pushing KVM on zCloud over just
> offering z/VM as a Service (zaas) for running Linux guest like they should
> be!
>
> On Mon, Oct 21, 2019 at 9:39 PM Barton 
> wrote:
>
> > Right.  That’s what I’m hearing , that ‘unlimited ‘ is not available on Z
> >
> > Barton
> >
> >
> > > On Oct 22, 2019, at 12:27 AM, Jim Elliott 
> wrote:
> > >
> > > Sam:
> > >
> > > The Red Hat site above does not make sense to me as there is only ONE
> > > version of RHEL available for IBM Z. That is, Red Hat Enterprise Linux
> > with
> > > Unlimited Guests and Red Hat Enterprise Virtualization are NOT
> available
> > > for IBM Z.
> > >
> > > Jim Elliott
> > > Senior Consultant - GlassHouse Systems Inc.
> > >
> > >
> > >
> > >> On Mon, Oct 21, 2019 at 6:07 PM Cohen, Sam  wrote:
> > >>
> > >> Barton,
> > >>
> > >> It depends on the licensed version of Redhat that you have.  From the
> > >> footnote:
> > >>
> > >> This guest limit does not apply to Red Hat Enterprise Linux with
> > Unlimited
> > >> Guests. There is no guest limit for Red Hat Enterprise Virtualization,
> > Red
> > >> Hat Enterprise Linux OpenStack Platform, and the Red Hat Enterprise
> > Linux
> > >> with Smart Virtualization bundle.
> > >>
> > >> Thanks,
> > >>
> > >> Sam
> > >> (217) 862-9227 (office)
> > >> (602) 327-2134 (cell)
> > >>
> > >> -Original Message-
> > >> From: Linux on 390 Port  On Behalf Of Barton
> > >> Sent: Monday, October 21, 2019 2:54 PM
> > >> To: LINUX-390@VM.MARIST.EDU
> > >> Subject: KVM only supports 4guests
> > >>
> > >> I’m hearing at gse that red hat on z only supports 4 guests on kvm on
> z?
> > >> Here’s a  link, am i missIng something?
> > >>
> >
> https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Faccess.redhat.com%2Farticles%2Frhel-kvm-limitsdata=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=g%2FPSoSxYqh2SgqhLX5V5KrvdkgFPSuWcvqrkupgaKj0%3Dreserved=0
> > >>
> > >> Barton
> > >>
> > >>
> > >> --
> > >> For LINUX-390 subscribe / signoff / archive access instructions, send
> > >> email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> > visit
> > >>
> > >>
> >
> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww2.marist.edu%2Fhtbin%2Fwlvindex%3FLINUX-390data=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=imwx9PkkgF74LtoqU5u9d2mv2nZ0HgHu5OXuUlEwqVo%3Dreserved=0
> > >>
> > >> --
> > >> For LINUX-390 subscribe / signoff / archive access instructions,
> > >> send email to lists...@vm.marist.edu with the message: INFO LINUX-390
> > or
> > >> visit
> > >> http://www2.marist.edu/htbin/wlvindex?LINUX-390
> > >>
> > >
> > > --
> > > For LINUX-390 subscribe / signoff / archive access instructions,
> > > send email to lists...@vm.marist.edu with the message: INFO LINUX-390
> > or visit
> > > http://www2.marist.edu/htbin/wlvindex?LINUX-390
> >
> > --
> > For LINUX-390 subscribe / signoff / archive access instructions,
> > send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> > visit
> > http://www2.marist.edu/htbin/wlvindex?LINUX-390
> >
>
>
> --
>
> ><º>`·.¸¸´¯`·.¸.·´¯`·...¸>(((º>
> .·´¯`·.><º>`·.¸¸.·´¯`·.¸.·´¯`·...¸><º>
>
> <>< Go fishing ><>
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390
>


-- 
Jay Brenneman

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM only supports 4guests

[Bfishing]

I wonder if that winds up depending on where you or they run it?
- I stated it that way knowing that IBM is pushing KVM on zCloud over just
offering z/VM as a Service (zaas) for running Linux guest like they should
be!

On Mon, Oct 21, 2019 at 9:39 PM Barton  wrote:

> Right.  That’s what I’m hearing , that ‘unlimited ‘ is not available on Z
>
> Barton
>
>
> > On Oct 22, 2019, at 12:27 AM, Jim Elliott  wrote:
> >
> > Sam:
> >
> > The Red Hat site above does not make sense to me as there is only ONE
> > version of RHEL available for IBM Z. That is, Red Hat Enterprise Linux
> with
> > Unlimited Guests and Red Hat Enterprise Virtualization are NOT available
> > for IBM Z.
> >
> > Jim Elliott
> > Senior Consultant - GlassHouse Systems Inc.
> >
> >
> >
> >> On Mon, Oct 21, 2019 at 6:07 PM Cohen, Sam  wrote:
> >>
> >> Barton,
> >>
> >> It depends on the licensed version of Redhat that you have.  From the
> >> footnote:
> >>
> >> This guest limit does not apply to Red Hat Enterprise Linux with
> Unlimited
> >> Guests. There is no guest limit for Red Hat Enterprise Virtualization,
> Red
> >> Hat Enterprise Linux OpenStack Platform, and the Red Hat Enterprise
> Linux
> >> with Smart Virtualization bundle.
> >>
> >> Thanks,
> >>
> >> Sam
> >> (217) 862-9227 (office)
> >> (602) 327-2134 (cell)
> >>
> >> -Original Message-
> >> From: Linux on 390 Port  On Behalf Of Barton
> >> Sent: Monday, October 21, 2019 2:54 PM
> >> To: LINUX-390@VM.MARIST.EDU
> >> Subject: KVM only supports 4guests
> >>
> >> I’m hearing at gse that red hat on z only supports 4 guests on kvm on z?
> >> Here’s a  link, am i missIng something?
> >>
> https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Faccess.redhat.com%2Farticles%2Frhel-kvm-limitsdata=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=g%2FPSoSxYqh2SgqhLX5V5KrvdkgFPSuWcvqrkupgaKj0%3Dreserved=0
> >>
> >> Barton
> >>
> >>
> >> --
> >> For LINUX-390 subscribe / signoff / archive access instructions, send
> >> email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> >>
> >>
> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww2.marist.edu%2Fhtbin%2Fwlvindex%3FLINUX-390data=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=imwx9PkkgF74LtoqU5u9d2mv2nZ0HgHu5OXuUlEwqVo%3Dreserved=0
> >>
> >> --
> >> For LINUX-390 subscribe / signoff / archive access instructions,
> >> send email to lists...@vm.marist.edu with the message: INFO LINUX-390
> or
> >> visit
> >> http://www2.marist.edu/htbin/wlvindex?LINUX-390
> >>
> >
> > --
> > For LINUX-390 subscribe / signoff / archive access instructions,
> > send email to lists...@vm.marist.edu with the message: INFO LINUX-390
> or visit
> > http://www2.marist.edu/htbin/wlvindex?LINUX-390
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390
>


-- 

><º>`·.¸¸´¯`·.¸.·´¯`·...¸>(((º>
.·´¯`·.><º>`·.¸¸.·´¯`·.¸.·´¯`·...¸><º>

<>< Go fishing ><>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM only supports 4guests

[Barton]

Right.  That’s what I’m hearing , that ‘unlimited ‘ is not available on Z

Barton


> On Oct 22, 2019, at 12:27 AM, Jim Elliott  wrote:
> 
> Sam:
> 
> The Red Hat site above does not make sense to me as there is only ONE
> version of RHEL available for IBM Z. That is, Red Hat Enterprise Linux with
> Unlimited Guests and Red Hat Enterprise Virtualization are NOT available
> for IBM Z.
> 
> Jim Elliott
> Senior Consultant - GlassHouse Systems Inc.
> 
> 
> 
>> On Mon, Oct 21, 2019 at 6:07 PM Cohen, Sam  wrote:
>> 
>> Barton,
>> 
>> It depends on the licensed version of Redhat that you have.  From the
>> footnote:
>> 
>> This guest limit does not apply to Red Hat Enterprise Linux with Unlimited
>> Guests. There is no guest limit for Red Hat Enterprise Virtualization, Red
>> Hat Enterprise Linux OpenStack Platform, and the Red Hat Enterprise Linux
>> with Smart Virtualization bundle.
>> 
>> Thanks,
>> 
>> Sam
>> (217) 862-9227 (office)
>> (602) 327-2134 (cell)
>> 
>> -Original Message-
>> From: Linux on 390 Port  On Behalf Of Barton
>> Sent: Monday, October 21, 2019 2:54 PM
>> To: LINUX-390@VM.MARIST.EDU
>> Subject: KVM only supports 4guests
>> 
>> I’m hearing at gse that red hat on z only supports 4 guests on kvm on z?
>> Here’s a  link, am i missIng something?
>> https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Faccess.redhat.com%2Farticles%2Frhel-kvm-limitsdata=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=g%2FPSoSxYqh2SgqhLX5V5KrvdkgFPSuWcvqrkupgaKj0%3Dreserved=0
>> 
>> Barton
>> 
>> 
>> --
>> For LINUX-390 subscribe / signoff / archive access instructions, send
>> email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
>> 
>> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww2.marist.edu%2Fhtbin%2Fwlvindex%3FLINUX-390data=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=imwx9PkkgF74LtoqU5u9d2mv2nZ0HgHu5OXuUlEwqVo%3Dreserved=0
>> 
>> --
>> For LINUX-390 subscribe / signoff / archive access instructions,
>> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
>> visit
>> http://www2.marist.edu/htbin/wlvindex?LINUX-390
>> 
> 
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM only supports 4guests

[Jim Elliott]

Sam:

The Red Hat site above does not make sense to me as there is only ONE
version of RHEL available for IBM Z. That is, Red Hat Enterprise Linux with
Unlimited Guests and Red Hat Enterprise Virtualization are NOT available
for IBM Z.

Jim Elliott
Senior Consultant - GlassHouse Systems Inc.



On Mon, Oct 21, 2019 at 6:07 PM Cohen, Sam  wrote:

> Barton,
>
> It depends on the licensed version of Redhat that you have.  From the
> footnote:
>
> This guest limit does not apply to Red Hat Enterprise Linux with Unlimited
> Guests. There is no guest limit for Red Hat Enterprise Virtualization, Red
> Hat Enterprise Linux OpenStack Platform, and the Red Hat Enterprise Linux
> with Smart Virtualization bundle.
>
> Thanks,
>
> Sam
> (217) 862-9227 (office)
> (602) 327-2134 (cell)
>
> -Original Message-
> From: Linux on 390 Port  On Behalf Of Barton
> Sent: Monday, October 21, 2019 2:54 PM
> To: LINUX-390@VM.MARIST.EDU
> Subject: KVM only supports 4guests
>
> I’m hearing at gse that red hat on z only supports 4 guests on kvm on z?
> Here’s a  link, am i missIng something?
> https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Faccess.redhat.com%2Farticles%2Frhel-kvm-limitsdata=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=g%2FPSoSxYqh2SgqhLX5V5KrvdkgFPSuWcvqrkupgaKj0%3Dreserved=0
>
> Barton
>
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions, send
> email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
>
> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww2.marist.edu%2Fhtbin%2Fwlvindex%3FLINUX-390data=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=imwx9PkkgF74LtoqU5u9d2mv2nZ0HgHu5OXuUlEwqVo%3Dreserved=0
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390
>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM only supports 4guests

[Mark Post]

On 10/21/19 5:53 PM, Barton wrote:
> I’m hearing at gse that red hat on z only supports 4 guests on kvm on z? 
> Here’s a  link, am i missIng something?  
> https://access.redhat.com/articles/rhel-kvm-limits

Yes. That's a licensing restriction, not a technical/performance one.
There's a superscript "1" by that number, and the corresponding footnote
says:
This guest limit does not apply to Red Hat Enterprise Linux with
Unlimited Guests. There is no guest limit for Red Hat Enterprise
Virtualization, Red Hat Enterprise Linux OpenStack Platform, and the Red
Hat Enterprise Linux with Smart Virtualization bundle.


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM only supports 4guests

[Cohen, Sam]

Barton,

It depends on the licensed version of Redhat that you have.  From the footnote:

This guest limit does not apply to Red Hat Enterprise Linux with Unlimited 
Guests. There is no guest limit for Red Hat Enterprise Virtualization, Red Hat 
Enterprise Linux OpenStack Platform, and the Red Hat Enterprise Linux with 
Smart Virtualization bundle.

Thanks,

Sam
(217) 862-9227 (office)
(602) 327-2134 (cell)

-Original Message-
From: Linux on 390 Port  On Behalf Of Barton
Sent: Monday, October 21, 2019 2:54 PM
To: LINUX-390@VM.MARIST.EDU
Subject: KVM only supports 4guests

I’m hearing at gse that red hat on z only supports 4 guests on kvm on z? Here’s 
a  link, am i missIng something?  
https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Faccess.redhat.com%2Farticles%2Frhel-kvm-limitsdata=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=g%2FPSoSxYqh2SgqhLX5V5KrvdkgFPSuWcvqrkupgaKj0%3Dreserved=0

Barton


--
For LINUX-390 subscribe / signoff / archive access instructions, send email to 
lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww2.marist.edu%2Fhtbin%2Fwlvindex%3FLINUX-390data=02%7C01%7CSam.Cohen%40LRS.COM%7C48d4b88bb27041ea857c08d75671af04%7C62af9ccc42164ae2a1d306614c59c315%7C0%7C0%7C637072918605890971sdata=imwx9PkkgF74LtoqU5u9d2mv2nZ0HgHu5OXuUlEwqVo%3Dreserved=0

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM and FCP disk without SAN switch

[Robert J Brenneman]

aahhh yea. a dark corner of memory gets its first ray of light
in a while

V7000 and other storwize related products ( SVC, FS9xxx ) now do NPIV on
the storage side by default if you are doing a new install. You have to
explicitly enable it if you are upgrading from older versions.

Ostensibly this feature is there to enable easier storage migrations down
the line, since the WWPN you're configuring into the SAN zone and on the
operating system side is now virtual. It is supposed to make switching the
physical storage box out from under the running OS by using box managed
copy services like metro-mirror easier and more automate-able without
having to take a planned outage.

This obviously also requires a SAN switch for it to work since it uses NPIV.

--
Jay Brenneman

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM and FCP disk without SAN switch

[Christian Borntraeger]

On 17.09.19 17:18, Johan Schelling wrote:
> I remember that when we started with KVM on LinuxONE around 3/4 years ago we 
> connected our V7000 directly (without a SAN switch) to the LinuxONE.  zVM 
> worked like a charm, but we had a lot of problems with (the then IBM 
> supplied) KVM…..  We had a lot of discussion with both the LinuxONE and the 
> storage guys as the documentation wasn’t quite clear on what was supported or 
> not. Problems we faced had to do with (hope I recall correctly) with WWPN’s 
> returned from the V7000 were  all mixed up… At the moment we added a SAN 
> switch all our problems were gone … ;-)
> Haven’t had any problems with KVM since …..   but haven’t checked the 
> situation without a SAN switch in the last years.

Yes, there was a firmware issue with the V7000s back then. But this has been 
fixed.
I still have a v7000 directly connected to my test z13 as an outcome of that 
bug back then.
But as I said, unless this is officially supported by all components you should 
better 
use a switch. 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM and FCP disk without SAN switch

[Johan Schelling]

I remember that when we started with KVM on LinuxONE around 3/4 years ago we 
connected our V7000 directly (without a SAN switch) to the LinuxONE.  zVM 
worked like a charm, but we had a lot of problems with (the then IBM supplied) 
KVM…..  We had a lot of discussion with both the LinuxONE and the storage guys 
as the documentation wasn’t quite clear on what was supported or not. Problems 
we faced had to do with (hope I recall correctly) with WWPN’s returned from the 
V7000 were  all mixed up… At the moment we added a SAN switch all our 
problems were gone … ;-)
Haven’t had any problems with KVM since …..   but haven’t checked the situation 
without a SAN switch in the last years.

Regards
Johan Schelling
Infrastructure Solution Architect


[cid:image001.png@01D008CB.2EEA8890]



ICU IT Services BV
Transistorstraat 55b  I  1322 CK  ALMERE

M 06 – 21 245 992  I  E 
johan.schell...@icu-it.nl
T 088 – 5 234 123  I  www.icu-it.nl  I  KvK 32135776
ICU Proclaimer

Op 17 sep. 2019, om 12:56 heeft Christian Borntraeger 
mailto:borntrae...@linux.ibm.com>> het volgende 
geschreven:

On 16.09.19 21:19, Jim Elliott wrote:
For z/VM you must have a SAN switch to connect FCP attached disk. Is this
also true for KVM (Ubuntu if it matters)?

As KVM borrows the FCP support from Linux the answer is the same as for Linux 
in LPAR.
In general the Linux code supports switched fabric and direct attachment. The 
code does
not support arbitrated loop.
What you also need is a support statement that tells about the storage server 
and the
Z system. For example a V7000 DOES work when connected to a z Box (there is no 
fencing)
but it is not a supported config (at least it was not when I check 3 years ago).
So in general you are on the safe-side with a switch. Without a switch requires 
to check
the certifications.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the 
message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM and FCP disk without SAN switch

[Robert J Brenneman]

Also be aware that if you want to use NPIV you /must/ have a switch and the
switch must explicitly support NPIV since that is a function that requires
both FCP Host adapter and switch support to work.

On Tue, Sep 17, 2019 at 7:35 AM Jim Elliott  wrote:

> Christian:
>
> Thanks. We will go with a switch to attach the V5010 and decided on an
> IBM/Cisco SAN32C-6 with 8-port 16 Gb bundle.
>
> Jim Elliott
> Senior Consultant - GlassHouse Systems Inc.
>
>
> On Tue, Sep 17, 2019 at 6:58 AM Christian Borntraeger <
> borntrae...@linux.ibm.com> wrote:
>
> > On 16.09.19 21:19, Jim Elliott wrote:
> > > For z/VM you must have a SAN switch to connect FCP attached disk. Is
> this
> > > also true for KVM (Ubuntu if it matters)?
> >
> > As KVM borrows the FCP support from Linux the answer is the same as for
> > Linux in LPAR.
> > In general the Linux code supports switched fabric and direct attachment.
> > The code does
> > not support arbitrated loop.
> > What you also need is a support statement that tells about the storage
> > server and the
> > Z system. For example a V7000 DOES work when connected to a z Box (there
> > is no fencing)
> > but it is not a supported config (at least it was not when I check 3
> years
> > ago).
> > So in general you are on the safe-side with a switch. Without a switch
> > requires to check
> > the certifications.
> >
> > --
> > For LINUX-390 subscribe / signoff / archive access instructions,
> > send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> > visit
> > http://www2.marist.edu/htbin/wlvindex?LINUX-390
> >
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390
>


--
Jay Brenneman

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM and FCP disk without SAN switch

[Jim Elliott]

Christian:

Thanks. We will go with a switch to attach the V5010 and decided on an
IBM/Cisco SAN32C-6 with 8-port 16 Gb bundle.

Jim Elliott
Senior Consultant - GlassHouse Systems Inc.


On Tue, Sep 17, 2019 at 6:58 AM Christian Borntraeger <
borntrae...@linux.ibm.com> wrote:

> On 16.09.19 21:19, Jim Elliott wrote:
> > For z/VM you must have a SAN switch to connect FCP attached disk. Is this
> > also true for KVM (Ubuntu if it matters)?
>
> As KVM borrows the FCP support from Linux the answer is the same as for
> Linux in LPAR.
> In general the Linux code supports switched fabric and direct attachment.
> The code does
> not support arbitrated loop.
> What you also need is a support statement that tells about the storage
> server and the
> Z system. For example a V7000 DOES work when connected to a z Box (there
> is no fencing)
> but it is not a supported config (at least it was not when I check 3 years
> ago).
> So in general you are on the safe-side with a switch. Without a switch
> requires to check
> the certifications.
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390
>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM and FCP disk without SAN switch

[Christian Borntraeger]

On 16.09.19 21:19, Jim Elliott wrote:
> For z/VM you must have a SAN switch to connect FCP attached disk. Is this
> also true for KVM (Ubuntu if it matters)?

As KVM borrows the FCP support from Linux the answer is the same as for Linux 
in LPAR.
In general the Linux code supports switched fabric and direct attachment. The 
code does
not support arbitrated loop.
What you also need is a support statement that tells about the storage server 
and the
Z system. For example a V7000 DOES work when connected to a z Box (there is no 
fencing)
but it is not a supported config (at least it was not when I check 3 years ago).
So in general you are on the safe-side with a switch. Without a switch requires 
to check
the certifications.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM and FCP disk without SAN switch

[Martha McConaghy]

Carlos is right.  It really doesn't matter what system is on the Z (VM 
or Linux), it has more to do with the FCP FICON adapters. The v 
boxes don't have the protocol to talk to FCP directly, while the more 
sophistcated DS8xxx do.  So, no matter what system you are talking to on 
Z, you would need a switch for the v5010.


Martha

On 9/16/2019 4:44 PM, Jim Elliott wrote:

Carlos:
  
Thanks. It will be a V5010 so we will configure a small SAN switch.
  
Jim Elliott

Senior Consultant - GlassHouse Systems Inc.
  
  
On Mon, Sep 16, 2019 at 4:39 PM Bodra - Pessoal  wrote:
  

 From our experience with KVM running on a zBox if you use as storage
devices IBM V5000 or V7000 you need to use a San Switch (2498-Bxx) to
connect. If you use storage devices IBM DS8800/8870, you can connect
directly. These are some tests that we did here at our shop.


Carlos Bodra
IBM zEnterprise Certified
São Paulo – SP – Brazil


-Mensagem original-
De: Linux on 390 Port  Em nome de Jim Elliott
Enviada em: segunda-feira, 16 de setembro de 2019 16:20
Para: LINUX-390@VM.MARIST.EDU
Assunto: KVM and FCP disk without SAN switch

For z/VM you must have a SAN switch to connect FCP attached disk. Is this
also true for KVM (Ubuntu if it matters)?

Jim Elliott
Senior Consultant - GlassHouse Systems Inc.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

  
--

For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390


--
Martha McConaghy
Marist: System Architect/Technical Lead
SHARE Association: Vice President
Marist College IT
Poughkeepsie, NY 12601

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM and FCP disk without SAN switch

[Jim Elliott]

Carlos:

Thanks. It will be a V5010 so we will configure a small SAN switch.

Jim Elliott
Senior Consultant - GlassHouse Systems Inc.


On Mon, Sep 16, 2019 at 4:39 PM Bodra - Pessoal  wrote:

> From our experience with KVM running on a zBox if you use as storage
> devices IBM V5000 or V7000 you need to use a San Switch (2498-Bxx) to
> connect. If you use storage devices IBM DS8800/8870, you can connect
> directly. These are some tests that we did here at our shop.
>
>
> Carlos Bodra
> IBM zEnterprise Certified
> São Paulo – SP – Brazil
>
>
> -Mensagem original-
> De: Linux on 390 Port  Em nome de Jim Elliott
> Enviada em: segunda-feira, 16 de setembro de 2019 16:20
> Para: LINUX-390@VM.MARIST.EDU
> Assunto: KVM and FCP disk without SAN switch
>
> For z/VM you must have a SAN switch to connect FCP attached disk. Is this
> also true for KVM (Ubuntu if it matters)?
>
> Jim Elliott
> Senior Consultant - GlassHouse Systems Inc.
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390
>

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM and FCP disk without SAN switch

[Adolph Kahan]

Yes, it is. With KVM you have more switch options than with z/VM

Adolph Kahan | Senior Consultant
Tel: +1 (416) 229-2950 x304 | Mobile: +1 (416) 822-3329
  

-Original Message-
From: Linux on 390 Port  On Behalf Of Jim Elliott
Sent: Monday, September 16, 2019 3:20 PM
To: LINUX-390@VM.MARIST.EDU
Subject: KVM and FCP disk without SAN switch

For z/VM you must have a SAN switch to connect FCP attached disk. Is this also 
true for KVM (Ubuntu if it matters)?

Jim Elliott
Senior Consultant - GlassHouse Systems Inc.

--
For LINUX-390 subscribe / signoff / archive access instructions, send email to 
lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
https://can01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww2.marist.edu%2Fhtbin%2Fwlvindex%3FLINUX-390data=01%7C01%7Cakahan%40GHSYSTEMS.COM%7C3c38157fe09f4553da0d08d73adb28c2%7C1c932bc9c5af4366811d012126996da4%7C0sdata=DwHDjMM7P9M4Rbc0M4TBx%2BYYHt9KlLhRfp9ncaYFMmM%3Dreserved=0

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM guest networking under z/VM

[Alan Altmark]

On Tuesday, 01/08/2019 at 05:14 GMT, Alan Haff  
wrote:
> I've asked the owner of the z/VM host to make the change. I'll report 
back with
> the results.

I also suggest that you move from Guest LAN to a VSWITCH.  By forcing the 
traffic to a z/VM-resident gateway, you add latency and increase CPU 
utilization.  Firewalls and routers should be external to the Z box.

Alan Altmark

Senior Managing z/VM and Linux Consultant
IBM Systems Lab Services
IBM Z Delivery Practice
ibm.com/systems/services/labservices
office: 607.429.3323
mobile; 607.321.7556
alan_altm...@us.ibm.com
IBM Endicott


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM guest networking under z/VM

[Alan Haff]

>-Original Message-
>From: Linux on 390 Port [mailto:LINUX-390@VM.MARIST.EDU] On Behalf Of 
>Christian Borntraeger
>Sent: Tuesday, January 8, 2019 00:36
>To: LINUX-390@VM.MARIST.EDU
>Subject: Re: KVM guest networking under z/VM
.
.
.
>as already said, can you use layer2?

I've asked the owner of the z/VM host to make the change. I'll report back with 
the results.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM guest networking under z/VM

[Christian Borntraeger]

On 03.01.2019 19:32, Alan Haff wrote:
> I'm setting up a KVM-hosted guest in a SLES12 VM running under z/VM and I'm 
> not able to get networking going in the KVM guest.
> 
> Host: z/VM 6.4
> KVM host: SLES12SP4
> KVM guest: SLES12SP4
> 
> (The KVM host must run under z/VM; I don't have any control over that.)

For what its worth, this will have non-ideal performance, but it should work
anyway. Reason is that z/VM "emulates" the hardware virtualization (the
sie instruction) in that case. In fact, it actually tries to translate KVMs
sie control blocks into a shadow control block and runs the KVM guest as a
shadowed z/VM guest and fakes the result so that the kvm host "sees" what it 
would see on real hardware. It also has to shadow all page table actions of
the kvm host, so I strongly suggest to have enough memory (no paging) in the
KVM host to minimize the impact. Not sure about z/VM, but I would guess that
paging in z/VM also hurts as this also requires tear down of shadow page
tables - this is at least the case when running KVM under KVM.

We usually do not test this scenario, but I know that some
developers do have at least one development system with KVM under z/VM so
it gets some testing.

Not sure if you can talk about that, but I am interested in the use case.
I expect to have z/VM guests all over the place (in your case another z/VM 
guest)
or KVM only installations. Mixed mode only with different LPARs. Are there
interesting use cases that would suggest that we invest more time in making sure
that KVM runs well under z/VM?

[...]
> The macvtap0 device does get created when I start the KVM guest with 'virsh 
> start sles12':

as already said, can you use layer2?

Christian

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM guest networking under z/VM

[Christian Borntraeger]

On 08.01.2019 01:35, Robert J Brenneman wrote:
> KVM Host bridges require a L2 network interface with the 'bridge_role'
> attribute set on the OSA device supporting the bridge.
> ref: https://public.dhe.ibm.com/software/dw/linux390/docu/lhs4dd05.pdf
> chapter 14, section Layer 2 promiscuous mode  on p205.
> 
> But you're not using a L2 Vswitch - the 'q lan' response indicates it's a
> L3 : 'PERSISTENT RESTRICTED IP'
> if it had been a L2 network it would have said 'PERSISTENT RESTRICTED ETH'
> 
> 
> But wait - you're not trying to run a KVM Bridge , you're doing MACVTAP
> which attaches the whole eth interface to the KVM guest.

Well, not exactly. macvtap does not attach the whole eth, it registers one
mac address at that interface (via osa address table) and then passes along
the traffic for that mac address to the guest. Now this (as well as bridges
and openvswitch) always require a layer 2 interface.

If you really HAVE to use a layer3 interface only (e.g. a hipersocket that
connects to z/OS) you can do routing in the KVM host.

Examples of how to use that can be found in the paper
Exploiting HiperSockets in a KVM Environment Using IP Routing with Linux on Z - 
Results and Findings
https://www-03.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/WP102746



> If you have the option to create a L2 LAN instead, try that first and see
> if it works.
> 
> If you can't create a L2 LAN to use for the MACVTAP interface, I am not
> sure whether the L3 LAN you have will work.  L3 LAN doesn't do ARPing at
> all so I'm not surprised wireshark is showing you no response to the
> outgoing ARP.  The virtual machine doesn't interact with the MACVTAP using
> the qeth driver internally, so you can't really put it in L3 mode to make
> it work that way.
> 
> I hope you have the option to create a L2 Guest LAN or vswitch for the KVM
> host's interface - the only other option I can think of would be to
> dedicate a triplet of OSA devices ( Read/Write/Data ) directly to the KVM
> Host's VM Guest.
> 
> On Thu, Jan 3, 2019 at 5:22 PM Alan Haff  wrote:
> 
>> I'm setting up a KVM-hosted guest in a SLES12 VM running under z/VM and
>> I'm not able to get networking going in the KVM guest.
>>
>> Host: z/VM 6.4
>> KVM host: SLES12SP4
>> KVM guest: SLES12SP4
>>
>> (The KVM host must run under z/VM; I don't have any control over that.)
>>
>> The KVM guest's network is defined as:
>>
>> 
>>   
>>   
>> 
>>
>> which, after 'virsh define sles12.xml', becomes:
>>
>> 
>>   
>>   
>>   
>>   
>>   
>>   
>> 
>>
>> The macvtap0 device does get created when I start the KVM guest with
>> 'virsh start sles12':
>>
>> 4: macvtap0@eth0:  mtu 1492
>> qdisc pfifo_fast state UP mode DEFAULT group default qlen 500
>> link/ether 52:54:00:93:f4:ce brd ff:ff:ff:ff:ff:ff
>>
>> Running WireShark on the KVM host when I try to ping the default gateway
>> from within the KVM guest, I can see ARP packets on the macvtap device (and
>> the transmit packet count increases) but I never see any ARP replies coming
>> back:
>>
>> macvtap0  Link encap:Ethernet  HWaddr 52:54:00:93:F4:CE
>>   UP BROADCAST RUNNING PROMISC MULTICAST  MTU:1492  Metric:1
>>   RX packets:0 errors:0 dropped:0 overruns:0 frame:0
>>   TX packets:10 errors:0 dropped:0 overruns:0 carrier:0
>>   collisions:0 txqueuelen:500
>>   RX bytes:0 (0.0 b)  TX bytes:420 (420.0 b)
>>
>> I've confirmed that MACPROTECT is off:
>>
>> q lan
>> 00: LAN SYSTEM LANATTC  Type: QDIOConnected: 5Maxconn: INFINITE
>> 00:   PERSISTENT  RESTRICTEDIPAccounting: OFF
>> 00:   IPTimeout: 5 MAC Protection: OFF
>> 00:   Isolation Status: OFFVEPA Status: OFF
>> q vmlan
>> 00: VMLAN maintenance level:
>> 00:   Latest Service: VM65918
>> 00: VMLAN MAC address assignment:
>> 00:   System MAC Protection: OFF
>> 00:   MACADDR Prefix: 02DF02 USER Prefix: 02DF02
>> 00:   MACIDRANGE SYSTEM: 01-FF
>> 00:  USER:   00-00
>> 00: VMLAN default accounting status:
>> 00:   SYSTEM Accounting: OFF   USER Accounting: OFF
>> 00: VMLAN general activity:
>> 00:   PERSISTENT Limit: INFINITE   Current: 20
>> 00:   TRANSIENT  Limit: INFINITE   Current: 0
>> 00: Trace Pages: 8
>> 00: VMLAN Directory Network Authorization: ENABLED
>> 00: IVL Domain: None
>>
>>
>> Any ideas about what I could be missing?

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM guest networking under z/VM

[Robert J Brenneman]

KVM Host bridges require a L2 network interface with the 'bridge_role'
attribute set on the OSA device supporting the bridge.
ref: https://public.dhe.ibm.com/software/dw/linux390/docu/lhs4dd05.pdf
chapter 14, section Layer 2 promiscuous mode  on p205.

But you're not using a L2 Vswitch - the 'q lan' response indicates it's a
L3 : 'PERSISTENT RESTRICTED IP'
if it had been a L2 network it would have said 'PERSISTENT RESTRICTED ETH'


But wait - you're not trying to run a KVM Bridge , you're doing MACVTAP
which attaches the whole eth interface to the KVM guest.
If you have the option to create a L2 LAN instead, try that first and see
if it works.

If you can't create a L2 LAN to use for the MACVTAP interface, I am not
sure whether the L3 LAN you have will work.  L3 LAN doesn't do ARPing at
all so I'm not surprised wireshark is showing you no response to the
outgoing ARP.  The virtual machine doesn't interact with the MACVTAP using
the qeth driver internally, so you can't really put it in L3 mode to make
it work that way.

I hope you have the option to create a L2 Guest LAN or vswitch for the KVM
host's interface - the only other option I can think of would be to
dedicate a triplet of OSA devices ( Read/Write/Data ) directly to the KVM
Host's VM Guest.

On Thu, Jan 3, 2019 at 5:22 PM Alan Haff  wrote:

> I'm setting up a KVM-hosted guest in a SLES12 VM running under z/VM and
> I'm not able to get networking going in the KVM guest.
>
> Host: z/VM 6.4
> KVM host: SLES12SP4
> KVM guest: SLES12SP4
>
> (The KVM host must run under z/VM; I don't have any control over that.)
>
> The KVM guest's network is defined as:
>
> 
>   
>   
> 
>
> which, after 'virsh define sles12.xml', becomes:
>
> 
>   
>   
>   
>   
>   
>   
> 
>
> The macvtap0 device does get created when I start the KVM guest with
> 'virsh start sles12':
>
> 4: macvtap0@eth0:  mtu 1492
> qdisc pfifo_fast state UP mode DEFAULT group default qlen 500
> link/ether 52:54:00:93:f4:ce brd ff:ff:ff:ff:ff:ff
>
> Running WireShark on the KVM host when I try to ping the default gateway
> from within the KVM guest, I can see ARP packets on the macvtap device (and
> the transmit packet count increases) but I never see any ARP replies coming
> back:
>
> macvtap0  Link encap:Ethernet  HWaddr 52:54:00:93:F4:CE
>   UP BROADCAST RUNNING PROMISC MULTICAST  MTU:1492  Metric:1
>   RX packets:0 errors:0 dropped:0 overruns:0 frame:0
>   TX packets:10 errors:0 dropped:0 overruns:0 carrier:0
>   collisions:0 txqueuelen:500
>   RX bytes:0 (0.0 b)  TX bytes:420 (420.0 b)
>
> I've confirmed that MACPROTECT is off:
>
> q lan
> 00: LAN SYSTEM LANATTC  Type: QDIOConnected: 5Maxconn: INFINITE
> 00:   PERSISTENT  RESTRICTEDIPAccounting: OFF
> 00:   IPTimeout: 5 MAC Protection: OFF
> 00:   Isolation Status: OFFVEPA Status: OFF
> q vmlan
> 00: VMLAN maintenance level:
> 00:   Latest Service: VM65918
> 00: VMLAN MAC address assignment:
> 00:   System MAC Protection: OFF
> 00:   MACADDR Prefix: 02DF02 USER Prefix: 02DF02
> 00:   MACIDRANGE SYSTEM: 01-FF
> 00:  USER:   00-00
> 00: VMLAN default accounting status:
> 00:   SYSTEM Accounting: OFF   USER Accounting: OFF
> 00: VMLAN general activity:
> 00:   PERSISTENT Limit: INFINITE   Current: 20
> 00:   TRANSIENT  Limit: INFINITE   Current: 0
> 00: Trace Pages: 8
> 00: VMLAN Directory Network Authorization: ENABLED
> 00: IVL Domain: None
>
>
> Any ideas about what I could be missing?
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> http://www.marist.edu/htbin/wlvindex?LINUX-390
> --
> For more information on Linux on System z, visit
> http://wiki.linuxvm.org/
>


--
Jay Brenneman

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM guest networking under z/VM

[Jeffrey Barnard]

Alan,

To me, it looks like you have defined a z/VM Guest LAN (LANATTC) that is
layer 3 (00:   PERSISTENT  RESTRICTEDIP)
vs. Layer 2 (PERSISTENT  UNRESTRICTED  ETHERNET)

Is the gateway a z/VM TCPIP machine?
How is the LINK defined to z/VM TCP/IP (IP - later3 or ETHERNET - layer2)?

What are the IP addresses on the z/VM TCPIP machine and the KVM host and
guest machines?

Regards,
Jeff

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM guest networking under z/VM

[Alan Haff]

>-Original Message-
>From: Linux on 390 Port [mailto:LINUX-390@VM.MARIST.EDU] On Behalf Of Alan Haff
>Sent: Thursday, January 3, 2019 10:32
>To: LINUX-390@VM.MARIST.EDU
>Subject: KVM guest networking under z/VM
.
.
.
>Any ideas about what I could be missing?

More info...

In the KVM guest I can see broadcast packets coming from other systems on the 
same subnet. When I try to ping one of those other systems from within the KVM 
guest, the guest ARPs for a MAC address and never receives a response.

Side note: it seems odd to me that the guest would ARP for the MAC address. I'd 
have thought the address would've been in the guest's ARP cache. Dumping the 
ARP cache in the guest shows the HWaddress for the other systems as 
"(incomplete)". 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM on Z: Getting started

[Mike Friesenegger]

The "Migrating from KVM for IBM z Systems to KVM integrated with SUSE 
Linux Enterprise Server" document - 
https://www.suse.com/documentation/suse-best-practices/kvm-migration-ibmz-to-sles/data/kvm-migration-ibmz-to-sles.html 
- can be used a getting started guide to install and manage KVM 
instances using KVM in SLES.


One item that is not included in the document is that I have also been 
using virt-manager running on my x86 openSUSE systems and connected to 
libvirtd on SLES on Z to create, manage and modify KVM Z instances.


Regards,

Mike
--
Mike Friesenegger
Technology Strategist - IBM Global Alliance
SUSE
mi...@suse.com
(P)+1 303.249.0817

On 12/01/2017 06:57 AM, Stefan Raspl wrote:

There's a new series of articles coming up providing step-by-step
instructions on how to run KVM on Z. See
http://kvmonz.blogspot.de/p/getting-started-with-kvm-on-z.html for
an overview, and http://kvmonz.blogspot.de/p/getting-started.html
for a first article covering SLES12 SP3.


Mit freundlichen Grüßen / Kind regards

Stefan Raspl


KVM & Linux on z Systems
IBM Systems & Technology Group, Systems Software Development / SW
Linux on
z Systems Dev & Service
---
IBM Deutschland
Schoenaicher Str. 220
71032 Boeblingen
Phone: +49-7031-16-2177
E-Mail: stefan.ra...@de.ibm.com
---
IBM Deutschland Research & Development GmbH / Vorsitzende des
Aufsichtsrats: Martina Koederitz
Geschäftsführung: Dirk Wittkopp
Sitz der Gesellschaft: Böblingen / Registergericht: Amtsgericht
Stuttgart, HRB 243294

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/





--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Agblad Tore]

Hi, we are possibly about to change from active/passive OSA into link 
aggregation.
Mostly due to the non-z part of the network folk want's the same setup as x86.
We have not seen any real disadvantages, but in this discussion I might see it.
Is it better to use active/passive, unless network load requires both to work.

/Tore

 
Tore Agblad 
zOpen Teamleader
IT Services

Volvo Group Headquarters
Corporate Process  IT
E-mail: tore.agb...@volvo.com 
http://www.volvo.com/volvoit/global/en-gb/ 


-Original Message-
From: Linux on 390 Port [mailto:LINUX-390@VM.MARIST.EDU] On Behalf Of Alan 
Altmark
Sent: den 18 februari 2015 2:18
To: LINUX-390@VM.MARIST.EDU
Subject: Re: KVM in Linux for system z capabilities

On Tuesday, 02/17/2015 at 05:33 EST, Berthold Gunreben b...@suse.de wrote:
 Well, I don't see it as an advantage over real link aggregation, but
 IMHO it is a real advantage over an active/passive failover.

It's worse.  Keep those packets in order!

 You might argue that LACP is available to z/VM, but the downside is,
 that it subverses the virtualization of the OSA adapters, because it
 reserves the port.

 Or is there a real link aggregation available in z/VM that works
 without a port group? Means, is there s link aggregation in z/VM that
 works on shared OSA adapters?

From the z13 announcement:

z/VM Multi-VSwitch Link Aggregation Support
With the PTFs for APARs VM65583 and PI21053, z/VM V6.3 provides
Multi-VSwitch Link Aggregation Support, allowing a port group of
OSA-Express features to span multiple virtual switches within a single
z/VM system or between multiple z/VM systems. Sharing a Link Aggregation
Port Group (LAG) with multiple virtual switches increases optimization and
utilization of the OSA-Express when handling larger traffic loads. Higher
adapter utilization protects customer investments, which is increasingly
important as 10 gigabit deployments become more prevalent. This
enhancement makes it possible to do VSwitch Link Aggregation with OSAs
shared with other z/VM logical partitions, lifting the previous
restriction of requiring dedicated OSAs.


 The advantage that I see here for Linux/KVM is, that you can run
 balance-rr on shared OSA devices. And while there is probably not much
 of a difference to the Mainframe side, the Network Switches are way
 more happy with this mode than with active/passive.


The network switch is neither happy nor sad.  :-)  As long as you don't
have excessive queuing or service times for the adapter, you don't need
more than one.

And unlike x86 devices, an OSA runs more smoothly when you keep data
running through it.  Start/Stop/Start/Stop just wastes time.


  And one related cautionary note:   I recently learned that you
  shouldn't put unbonded adapters on the same subnet without doing your
  homework.  By default, Linux will consolidate all traffic onto just
  one of those adapters, it having all of the same-subnet IP addresses
  consolidated onto it.  (Very odd when you see QUERY VSWITCH DETAILS
  output showing adapters with no IP addresses and others with more
  than one!)I'm not really sure *why* Linux does that, but I'm
  guessing that it's an x86 thing.   Or, it could just be to annoy me.

 Is it really linux that does this? I normally see, that without a port
 group, you have an active adapter, and a backup adapter for a VSWITCH.

Yes, it's really true.   If your routing table and the return path filter
(rp_filter) aren't in agreement, packets get discarded.   Linux moves the
IP addresses to a single adapter.  See
http://www.novell.com/support/kb/doc.php?id=7007649

Alan Altmark

Senior Managing z/VM and Linux Consultant
Lab Services System z Delivery Practice
IBM Systems  Technology Group
ibm.com/systems/services/labservices
office: 607.429.3323
mobile; 607.321.7556
alan_altm...@us.ibm.com
IBM Endicott

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Alan Altmark]

On Tuesday, 02/17/2015 at 05:33 EST, Berthold Gunreben b...@suse.de wrote:
 Well, I don't see it as an advantage over real link aggregation, but
 IMHO it is a real advantage over an active/passive failover.

It's worse.  Keep those packets in order!

 You might argue that LACP is available to z/VM, but the downside is,
 that it subverses the virtualization of the OSA adapters, because it
 reserves the port.

 Or is there a real link aggregation available in z/VM that works
 without a port group? Means, is there s link aggregation in z/VM that
 works on shared OSA adapters?

From the z13 announcement:

z/VM Multi-VSwitch Link Aggregation Support
With the PTFs for APARs VM65583 and PI21053, z/VM V6.3 provides
Multi-VSwitch Link Aggregation Support, allowing a port group of
OSA-Express features to span multiple virtual switches within a single
z/VM system or between multiple z/VM systems. Sharing a Link Aggregation
Port Group (LAG) with multiple virtual switches increases optimization and
utilization of the OSA-Express when handling larger traffic loads. Higher
adapter utilization protects customer investments, which is increasingly
important as 10 gigabit deployments become more prevalent. This
enhancement makes it possible to do VSwitch Link Aggregation with OSAs
shared with other z/VM logical partitions, lifting the previous
restriction of requiring dedicated OSAs.


 The advantage that I see here for Linux/KVM is, that you can run
 balance-rr on shared OSA devices. And while there is probably not much
 of a difference to the Mainframe side, the Network Switches are way
 more happy with this mode than with active/passive.


The network switch is neither happy nor sad.  :-)  As long as you don't
have excessive queuing or service times for the adapter, you don't need
more than one.

And unlike x86 devices, an OSA runs more smoothly when you keep data
running through it.  Start/Stop/Start/Stop just wastes time.


  And one related cautionary note:   I recently learned that you
  shouldn't put unbonded adapters on the same subnet without doing your
  homework.  By default, Linux will consolidate all traffic onto just
  one of those adapters, it having all of the same-subnet IP addresses
  consolidated onto it.  (Very odd when you see QUERY VSWITCH DETAILS
  output showing adapters with no IP addresses and others with more
  than one!)I'm not really sure *why* Linux does that, but I'm
  guessing that it's an x86 thing.   Or, it could just be to annoy me.

 Is it really linux that does this? I normally see, that without a port
 group, you have an active adapter, and a backup adapter for a VSWITCH.

Yes, it's really true.   If your routing table and the return path filter
(rp_filter) aren't in agreement, packets get discarded.   Linux moves the
IP addresses to a single adapter.  See
http://www.novell.com/support/kb/doc.php?id=7007649

Alan Altmark

Senior Managing z/VM and Linux Consultant
Lab Services System z Delivery Practice
IBM Systems  Technology Group
ibm.com/systems/services/labservices
office: 607.429.3323
mobile; 607.321.7556
alan_altm...@us.ibm.com
IBM Endicott

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Berthold Gunreben]

On Tue, 17 Feb 2015 20:18:00 -0500
Alan Altmark alan_altm...@us.ibm.com wrote:
 
 From the z13 announcement:
 
 z/VM Multi-VSwitch Link Aggregation Support
 With the PTFs for APARs VM65583 and PI21053, z/VM V6.3 provides
 Multi-VSwitch Link Aggregation Support, allowing a port group of
 OSA-Express features to span multiple virtual switches within a single
 z/VM system or between multiple z/VM systems. Sharing a Link
 Aggregation Port Group (LAG) with multiple virtual switches increases
 optimization and utilization of the OSA-Express when handling larger
 traffic loads. Higher adapter utilization protects customer
 investments, which is increasingly important as 10 gigabit
 deployments become more prevalent. This enhancement makes it possible
 to do VSwitch Link Aggregation with OSAs shared with other z/VM
 logical partitions, lifting the previous restriction of requiring
 dedicated OSAs.

Now, that is a real improvement. I wasn't aware of it. Unfortunately,
we do not yet have z13.
 
 Yes, it's really true.   If your routing table and the return path
 filter (rp_filter) aren't in agreement, packets get discarded.
 Linux moves the IP addresses to a single adapter.  See
 http://www.novell.com/support/kb/doc.php?id=7007649

Ahh .. now I understand what you are talking about. Yes, that is an
issue especially if you have a multihomed server. 

You actually can configure this correctly by using network ACLs. I
don't know if that is nicely documented though.

Berthold

-- 
--
 Berthold Gunreben  Build Service Team
 http://www.suse.de/ Maxfeldstr. 5
 SUSE LINUX GmbHD-90409 Nuernberg, Germany
 GF: Felix Imendörffer, Jane Smithard, Jennifer Guild, Dilip Upmanyu,
Graham Norton HRB 21284 (AG Nürnberg) 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Mark Post]

 On 2/18/2015 at 02:34 AM, Berthold Gunreben b...@suse.de wrote: 
 You actually can configure this correctly by using network ACLs. I
 don't know if that is nicely documented though.

It wasn't when i tried to figure it out a couple of years ago.  I don't think 
that's changed.


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Alan Altmark]

On Tuesday, 02/17/2015 at 07:10 EST, Berthold Gunreben b...@suse.de wrote:
 There are some advantages to KVM on System z too:
 - Network bond balance-rr to have shared network devices and high
 availability network (active-passive does not count as highly
 available network for certain external switches).

(cough) Why would you classify balance-rr as an 'advantage', Berthold?
That's a primitive per-packet (subject to packets_per_slave) round robin
distribution. a truly horrible idea since in an HA environment those two
adapters go to different switches.  All you're doing is increasing the
risk of out-of-order packet arrival at the destination, burning CPU time
to reassemble the TCP segments and delays segment delivery to the socket
(app).

At least use balance-xor so that the assignment to an adapter is made
based on MAC address and is persistent unless there's a failure.   Add
xmit_hash_policy layer3+4 and you have winner.  (You really want layer3+4,
but you can't use it with balance-rr.)

But these days there's really no excuse for not using IEEE 802.3ad.   It's
standard on a switch of any reasonable vintage, and these days they all
include support for multi-switch port groups (aka virtual chassis).

MVS has per-packet load balancing, too.  It was abandoned in favor of
per-TCP session distribution (layer3+4).

Since data queued on a virtual NIC is immediately copied into CP and the
guest QDIO queues cleared, multiple NICs to the same VSWITCH (and VLAN)
serve only to spend more CPU time getting the data to the same place (an
outbound OSA queue).   IEEE 802.3ad takes are of distributing that traffic
among the OSAs in the port group.

And one related cautionary note:   I recently learned that you shouldn't
put unbonded adapters on the same subnet without doing your homework.  By
default, Linux will consolidate all traffic onto just one of those
adapters, it having all of the same-subnet IP addresses consolidated onto
it.  (Very odd when you see QUERY VSWITCH DETAILS output showing adapters
with no IP addresses and others with more than one!)I'm not really
sure *why* Linux does that, but I'm guessing that it's an x86 thing.   Or,
it could just be to annoy me.

Alan Altmark

Senior Managing z/VM and Linux Consultant
Lab Services System z Delivery Practice
IBM Systems  Technology Group
ibm.com/systems/services/labservices
office: 607.429.3323
mobile; 607.321.7556
alan_altm...@us.ibm.com
IBM Endicott

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Mark Post]

 On 2/16/2015 at 09:19 PM, Phil Tully tull...@optonline.net wrote: 
 In my opinion( soap box alert),  KVM on Z architecture is a mistake 
 promulgated by inane market speak managers in IBM that don't see the damage 
 they are doing by diverting resources from the currently most capable 
 virtualization product one the market.   Currently, because continued lack 
 of development will allow the lead to slide away from z/VM and therefore the 
 Virtual Z platform.

Even managers in IBM don't get to control the open source community.  There 
have been mainframe people interested in KVM for a long time, and it was going 
to happen one way or the other.  Having IBM employees contribute was helpful in 
terms of not having KVM on System z stink in comparison to the other 
architectures.  And I seriously doubt that any of the Linux developers (who 
were already working on Linux for IBM) could be reassigned to work on z/VM and 
be effective very quickly.

So, while I share your belief that z/VM is the gold standard for virtualization 
today, I don't see KVM overtaking that any time soon.  I would add if ever, 
but that would just be tempting fate and history.


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Mark Post]

 On 2/17/2015 at 02:07 AM, Pavelka, Tomas tomas.pave...@ca.com wrote: 
 Could you elaborate on what are the most important things that z/VM provides 
 that KVM does not? I occasionally get questions from people that heard 
 somewhere that KVM will replace z/VM on the mainframe but can't give answers 
 because I don't really know anything about KVM.

Berthold gave a decent list of items.  I would add to that
- More dynamic capabilities, even for non-privileged guests.  I can't count the 
number of times I wanted to add a device to a KVM guest and had to wait until 
it was shutdown and restarted before I could use it.
- Better sharing capabilities
- Really good performance data collection
- Better isolation/separation of guests
- Better over-commit of resources
- CP Privilege classes
- External Security Manager (ESM) interface(s)
- DCSS
- xip2fs
- Named Saved Systems
- CP USER DIRECT is a lot easier for a human to parse than KVM's XML files.

But in large part, z/VM is better because of the ISV ecosystem around it.  
There are a huge number of management tools out there for z/VM (which can also 
be a problem for shops to figure out which are the ones they want).  Being so 
new, KVM really has nothing like that.  I like libvirt, but it also is not yet 
very mature.  VMware is the dominant hypervisor in the distributed world, and 
will likely remain so for some time.  A big part of that are the management 
tools that VMware creates.  In the mainframe world, z/VM is the same way, but 
from a multiplicity of vendors, not just IBM.

I'm sure other people could add a lot of items to this list, since I just 
rattled this off from the top of my head.  I like KVM and think it's cool.  
SUSE uses it for our internal Build Service for System z, which is for us a 
mission critical application.  I'm just not ready to recommend it to customers 
for _their_ business critical workload because it hasn't been used in the real 
world enough yet.


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Mark Post]

 On 2/17/2015 at 10:16 AM, Alan Altmark alan_altm...@us.ibm.com wrote: 
 I'm not really
 sure *why* Linux does that, but I'm guessing that it's an x86 thing.   Or,
 it could just be to annoy me.

I would go with the latter possibility.  :)


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Berthold Gunreben]

On Tue, 17 Feb 2015 10:16:25 -0500
Alan Altmark alan_altm...@us.ibm.com wrote:

 On Tuesday, 02/17/2015 at 07:10 EST, Berthold Gunreben b...@suse.de
 wrote:
  There are some advantages to KVM on System z too:
  - Network bond balance-rr to have shared network devices and high
  availability network (active-passive does not count as highly
  available network for certain external switches).
 
 (cough) Why would you classify balance-rr as an 'advantage', Berthold?

Well, I don't see it as an advantage over real link aggregation, but
IMHO it is a real advantage over an active/passive failover.

You might argue that LACP is available to z/VM, but the downside is,
that it subverses the virtualization of the OSA adapters, because it
reserves the port.

Or is there a real link aggregation available in z/VM that works
without a port group? Means, is there s link aggregation in z/VM that
works on shared OSA adapters?

I am yet to read about such a possibility, but then, you are the real
expert in that area.

The advantage that I see here for Linux/KVM is, that you can run
balance-rr on shared OSA devices. And while there is probably not much
of a difference to the Mainframe side, the Network Switches are way
more happy with this mode than with active/passive.

...
 And one related cautionary note:   I recently learned that you
 shouldn't put unbonded adapters on the same subnet without doing your
 homework.  By default, Linux will consolidate all traffic onto just
 one of those adapters, it having all of the same-subnet IP addresses
 consolidated onto it.  (Very odd when you see QUERY VSWITCH DETAILS
 output showing adapters with no IP addresses and others with more
 than one!)I'm not really sure *why* Linux does that, but I'm
 guessing that it's an x86 thing.   Or, it could just be to annoy me.

Is it really linux that does this? I normally see, that without a port
group, you have an active adapter, and a backup adapter for a VSWITCH.

You can easily differentiate between those adapters within linux, and
it is also not difficult to add different IP addresses to different
adapters, but it is the VSWITCH that takes care to which OSA device the
traffice is going.


Berthold

-- 
--
 Berthold Gunreben  Build Service Team
 http://www.suse.de/ Maxfeldstr. 5
 SUSE LINUX GmbHD-90409 Nuernberg, Germany
 GF: Felix Imendörffer, Jane Smithard, Jennifer Guild, Dilip Upmanyu,
Graham Norton HRB 21284 (AG Nürnberg) 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Berthold Gunreben]

On Tue, 17 Feb 2015 07:07:47 +
Pavelka, Tomas tomas.pave...@ca.com wrote:

  I envision people using KVM on System z eventually deciding they
  want to upgrade to z/VM for the better capabilities and
  manageability it provides.
 
 Could you elaborate on what are the most important things that z/VM
 provides that KVM does not? I occasionally get questions from people
 that heard somewhere that KVM will replace z/VM on the mainframe but
 can't give answers because I don't really know anything about KVM.

From my point of view, there are really some advantages for z/VM,
namely:

- Guests have much more possibilities to modify and manage their
  virtual machines. 
- Console access without having to grant access to the Hypervisor
  (both, define graf and iucv consoles)
- start, stop, redeployment, IUCV by the guest, similar things would
  need root access on KVM host.
- nice options to manage memory like swap on VDISK, thus needs less
  real memory.
- Dedicated disk and network devices, retaining the ability to live
  relocate VMs 
- Split z/VM management by tasks (not just one root for everything)
- virtual switch included


There are some advantages to KVM on System z too:
- Any knowledgeable Linux admin will be able to administer KVM on z.
- Network bond balance-rr to have shared network devices and high
  availability network (active-passive does not count as highly
  available network for certain external switches).
- update management for Linux and thus KVM is IMHO better
Disadvantage KVM:
- no direct attached devices to guest systems

For both:
- Mainframe is IMHO _the best_ virtualization platform. Penalty on disk
  and network IO is minimal compared to other platforms.
- The Mainframe can easily cope with system loads at 100% and still
  give reasonable performance for interactive and non-interactive use.
- IO subsystem is way beyond other platforms, no stalls of other
  machines just because one guest runs wild

Thats what just came to my mind, this list is likely not complete...

Berthold

-- 
--
 Berthold Gunreben  Build Service Team
 http://www.suse.de/ Maxfeldstr. 5
 SUSE LINUX GmbHD-90409 Nuernberg, Germany
 GF: Felix Imendörffer, Jane Smithard, Jennifer Guild, Dilip Upmanyu,
Graham Norton HRB 21284 (AG Nürnberg) 

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Keith Gooding]

Thanks Mark. That is more or less what I thought. BTW I only just noticed that 
the IBM-speak in the SOD in the z13 announcement said 'IBM intends to 
**support** a Kernel-based Virtual Machine (KVM) offering for z Systems' (my 
highlighting). ie 'support' not 'deliver'. 
Keith


 On Monday, 16 February 2015, 18:04, Mark Post mp...@suse.com wrote:
   

  On 2/16/2015 at 12:40 PM, Keith Gooding kw...@yahoo.co.uk wrote: 
 Can anyone help resolve this difference of opinion with colleagues ?
 
  I have always assumed from what I have read that the QEMU and KVM, offered 
 as a 'technology preview' in for instance SUSE 11 SP3 or SUSE 12 will only 
 run zLinux guests and its purpose is to provide an alternative to z/VM (some 
 day). Colleages say that it will emulate Intel so that you will be able to 
 run x386_64 Linux guests. Who is correct (maybe both ?).

While qemu is _capable_ of emulating other architectures, that's not how we 
build it or ship it.  The intent is not to provide an alternative to z/VM, 
however.  The intent is to help new-to-the-mainframe customers get started with 
virtualization on the platform using a tool they probably already know pretty 
well.  Lowering the barriers to entry in order to help convince them that 
mainframes really are a good idea.  I envision people using KVM on System z 
eventually deciding they want to upgrade to z/VM for the better capabilities 
and manageability it provides.

 Also does anyone have any guesses about IBMs Statement of Direction about 
 KVM that was in the z13 announcement. Will it be z13 only ?.  An IBM product 
 like z/VM ?. Will it emulate Intel ?

My pure, unadulterated, WAG is that it will be z13 (or later) only, a priced 
IBM product, and that it will not emulate Intel.  That just feels how IBM would 
approach things given their history.


Mark Post



--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Phil Tully]

In my opinion( soap box alert),  KVM on Z architecture is a mistake promulgated 
by inane market speak managers in IBM that don't see the damage they are doing 
by diverting resources from the currently most capable virtualization product 
one the market.   Currently, because continued lack of development will allow 
the lead to slide away from z/VM and therefore the Virtual Z platform.

Sent from my iPhone

 On Feb 16, 2015, at 1:36 PM, Keith Gooding kw...@yahoo.co.uk wrote:

 Thanks Mark. That is more or less what I thought. BTW I only just noticed 
 that the IBM-speak in the SOD in the z13 announcement said 'IBM intends to 
 **support** a Kernel-based Virtual Machine (KVM) offering for z Systems' (my 
 highlighting). ie 'support' not 'deliver'.
 Keith


 On Monday, 16 February 2015, 18:04, Mark Post mp...@suse.com wrote:


 On 2/16/2015 at 12:40 PM, Keith Gooding kw...@yahoo.co.uk wrote:
 Can anyone help resolve this difference of opinion with colleagues ?

   I have always assumed from what I have read that the QEMU and KVM, offered
 as a 'technology preview' in for instance SUSE 11 SP3 or SUSE 12 will only
 run zLinux guests and its purpose is to provide an alternative to z/VM (some
 day). Colleages say that it will emulate Intel so that you will be able to
 run x386_64 Linux guests. Who is correct (maybe both ?).

 While qemu is _capable_ of emulating other architectures, that's not how we 
 build it or ship it.  The intent is not to provide an alternative to z/VM, 
 however.  The intent is to help new-to-the-mainframe customers get started 
 with virtualization on the platform using a tool they probably already know 
 pretty well.  Lowering the barriers to entry in order to help convince them 
 that mainframes really are a good idea.  I envision people using KVM on 
 System z eventually deciding they want to upgrade to z/VM for the better 
 capabilities and manageability it provides.

 Also does anyone have any guesses about IBMs Statement of Direction about
 KVM that was in the z13 announcement. Will it be z13 only ?.  An IBM product
 like z/VM ?. Will it emulate Intel ?

 My pure, unadulterated, WAG is that it will be z13 (or later) only, a priced 
 IBM product, and that it will not emulate Intel.  That just feels how IBM 
 would approach things given their history.


 Mark Post



 --
 For LINUX-390 subscribe / signoff / archive access instructions,
 send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
 http://www.marist.edu/htbin/wlvindex?LINUX-390
 --
 For more information on Linux on System z, visit
 http://wiki.linuxvm.org/

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM in Linux for system z capabilities

[Pavelka, Tomas]

 I envision people using KVM on System z eventually deciding they want to 
 upgrade to z/VM for the better capabilities and manageability it provides.

Could you elaborate on what are the most important things that z/VM provides 
that KVM does not? I occasionally get questions from people that heard 
somewhere that KVM will replace z/VM on the mainframe but can't give answers 
because I don't really know anything about KVM.

Thanks,
Tomas

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM network setup on SUSE SLES

[Brad Gartside]

After I click OK, The dialog box goes away and I am then back to the
Virtual Machine Manager. It doesn't have anything defined and looks exactly
as it did when I started the Virtual Machine Manager.

Thank you
brad.garts...@gmail.com

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM network setup on SUSE SLES

[Brad Gartside]

I am still having trouble getting the first VM installed.

From the Virtual Machine Manager

I select Create a Virtual Machine


From the Install an Operating System? screen I select the following:

I need to install an operating system

I select SUSE Linux Enterprise Server 11

I setup the options as follows:

Hardware
Initial Memory1024MB
Maximum Memory2048MB
Virtual Processors1
Host DevicesNone
Peripheral Devices
Graphics AdapterNo Graphics Support
Keymap
Sound CardNone
Disks
1: 6.0 GB Hard Disk (file:/mnt/202/sles11/disk0.raw)
2: 3.1 GB CD-ROM or DVD (file:/mnt/203/sles11sp3.iso)
Network Adapters
1: QEMU Virtualzied NIC Card; Randomly generated MAC address
Operating System Installation
Operating SystemSUSE Linux Enterprise Sever 11
Installation Source3.1 GB CD-ROM or DVD
(file:/mnt/203/sles11sp3.iso)
Automated Installation
Additional Arguments


While it is installing I get the following error:

Error
internal error: process exited while connecting to monitor:
OK

Is there a log file created from this process? Has anyone else run into
this error?

Thank you
brad.garts...@gmail.com

On Wed, Jan 7, 2015 at 11:57 AM, Mark Post mp...@suse.com wrote:

  On 1/7/2015 at 11:30 AM, Brad Gartside brad.garts...@gmail.com
 wrote:
  The install process I am talking about is the yast menu system. I start
  yast from the ssh console or select it from the X window lower left hand
  start bar. A menu displays in the ssh console that is titled YaST2
  Control Center. In the left hand column I select Virtualization and in
  the right hand column I select Install Hypervisor and Tools. I am then
  asked Xen or KVM. I select KVM. It then shows a screen that lists the
 steps
  of the installation.
   Verify Installed Packages
   Network Bridge Configuration
  The Verify Installed Packages installs the following:
  libvirt
  libvirt-python
  vm-install
  kvm
  virt-manager
  virt-viewer
 
  The packages install without error. The process never asks to setup a
  Network Bridge Configuration. It then tells me
  For installing KVM guests, reboot the machine to load the necessary
  drivers., and then I am back to the main yast menu.

 With SLES12, that message is now incorrect.  For SLES11 SP3 it is
 necessary to reboot for things to work.  You won't be asked to configure a
 bridge at this point, and it's not needed, and won't work.

  Could you point me to documentation on a macvtap setup?

 The easiest way is just to use virt-manager to create a new KVM guest.  It
 should automatically create the macvtap interface for you.


 Mark Post

 --
 For LINUX-390 subscribe / signoff / archive access instructions,
 send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
 visit
 http://www.marist.edu/htbin/wlvindex?LINUX-390
 --
 For more information on Linux on System z, visit
 http://wiki.linuxvm.org/




--
Brad Gartside
4 Penzance Dr
Bella Vista, AR 72714
(479) 236-4355
brad.garts...@gmail.com

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM network setup on SUSE SLES

[Mark Post]

 On 1/9/2015 at 03:35 PM, Brad Gartside brad.garts...@gmail.com wrote: 
 Operating System Installation
 Operating SystemSUSE Linux Enterprise Sever 11
 Installation Source3.1 GB CD-ROM or DVD
 (file:/mnt/203/sles11sp3.iso)

Oh, and I would change the Installation Source to a network installation 
server.  I tend to use FTP since I can look at the FTP server logs if something 
doesn't seem to be working right.


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM network setup on SUSE SLES

[Mark Post]

 On 1/9/2015 at 03:35 PM, Brad Gartside brad.garts...@gmail.com wrote: 
 I am still having trouble getting the first VM installed.
-snip-
 While it is installing I get the following error:
 
 Error
 internal error: process exited while connecting to monitor:
 OK
 
 Is there a log file created from this process? Has anyone else run into
 this error?

Yes, there are, in various places.  Figuring out if one or more of them is 
relevant is the hard part.
~/.virt-manager/
/var/log/libvirt/

Do you have all available maintenance for SP3 installed on your host system?


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM network setup on SUSE SLES

[Brad Gartside]

The host system is SLES 12 GA

Thank you

On Jan 9, 2015, at 2:19 PM, Mark Post mp...@suse.com wrote:

 On 1/9/2015 at 03:35 PM, Brad Gartside brad.garts...@gmail.com wrote:
 I am still having trouble getting the first VM installed.
 -snip-
 While it is installing I get the following error:

Error
internal error: process exited while connecting to monitor:
OK

 Is there a log file created from this process? Has anyone else run into
 this error?

 Yes, there are, in various places.  Figuring out if one or more of them is 
 relevant is the hard part.
 ~/.virt-manager/
 /var/log/libvirt/

 Do you have all available maintenance for SP3 installed on your host system?


 Mark Post

 --
 For LINUX-390 subscribe / signoff / archive access instructions,
 send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
 http://www.marist.edu/htbin/wlvindex?LINUX-390
 --
 For more information on Linux on System z, visit
 http://wiki.linuxvm.org/

Rocket Software, Inc. and subsidiaries ■ 77 Fourth Avenue, Waltham MA 02451 ■ 
+1 800.966.3270 ■ +1 781.577.4321
Unsubscribe From Commercial Email – unsubscr...@rocketsoftware.com
Manage Your Subscription Preferences - 
http://info.rocketsoftware.com/GlobalSubscriptionManagementEmailFooter_SubscriptionCenter.html
Privacy Policy - http://www.rocketsoftware.com/company/legal/privacy-policy


This communication and any attachments may contain confidential information of 
Rocket Software, Inc. All unauthorized use, disclosure or distribution is 
prohibited. If you are not the intended recipient, please notify Rocket 
Software immediately and destroy all copies of this communication. Thank you.

Re: KVM network setup on SUSE SLES

[Mark Post]

 On 1/9/2015 at 04:20 PM, Brad Gartside bgarts...@rocketsoftware.com 
 wrote: 
 The host system is SLES 12 GA

The same question applies there.  Do you have all available maintenance 
installed on the host system?

Digging a bit deeper, once you click OK on that dialog box, does virt-manager 
show the guest still running and consuming CPU, etc?  Or does it go away 
completely?


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM network setup on SUSE SLES

[Brad Gartside]

The install process I am talking about is the yast menu system. I start
yast from the ssh console or select it from the X window lower left hand
start bar. A menu displays in the ssh console that is titled YaST2
Control Center. In the left hand column I select Virtualization and in
the right hand column I select Install Hypervisor and Tools. I am then
asked Xen or KVM. I select KVM. It then shows a screen that lists the steps
of the installation.
 Verify Installed Packages
 Network Bridge Configuration
The Verify Installed Packages installs the following:
libvirt
libvirt-python
vm-install
kvm
virt-manager
virt-viewer

The packages install without error. The process never asks to setup a
Network Bridge Configuration. It then tells me
For installing KVM guests, reboot the machine to load the necessary
drivers., and then I am back to the main yast menu.

Could you point me to documentation on a macvtap setup?

Thank you
brad.garts...@gmail.com

On Tue, Jan 6, 2015 at 11:26 PM, Mark Post mp...@suse.com wrote:

  On 1/6/2015 at 05:03 PM, Brad Gartside brad.garts...@gmail.com
 wrote:
  I am setting up a KVM proof of concept system on a z/Linux guest running
 on
  z/VM. I have tried using the GUI and text versions of yast to Install
  Hypervisor and Tools, and the install process always skips over the
  Network Bridge Configuration setup step.

 Which install process are you talking about here?

  The behavior has been repeated
  on SLES 11 SP3 GA and SLES 12 GA. The z/Linux KVM host guest has a Layer
 2
  OSA networking defined.
 
  Has anyone else run into network problems setting up KVM on SLES?

 Not a problem, exactly, but you need to be aware that using macvtap is the
 only thing that is going to work for you, due to the nature of the OSA
 hardware itself.  The virt-manager interface should set that up for you
 automatically, however.


 Mark Post

 --
 For LINUX-390 subscribe / signoff / archive access instructions,
 send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
 visit
 http://www.marist.edu/htbin/wlvindex?LINUX-390
 --
 For more information on Linux on System z, visit
 http://wiki.linuxvm.org/




--
Brad Gartside
4 Penzance Dr
Bella Vista, AR 72714
(479) 236-4355
brad.garts...@gmail.com

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM network setup on SUSE SLES

[Mark Post]

 On 1/7/2015 at 11:30 AM, Brad Gartside brad.garts...@gmail.com wrote: 
 The install process I am talking about is the yast menu system. I start
 yast from the ssh console or select it from the X window lower left hand
 start bar. A menu displays in the ssh console that is titled YaST2
 Control Center. In the left hand column I select Virtualization and in
 the right hand column I select Install Hypervisor and Tools. I am then
 asked Xen or KVM. I select KVM. It then shows a screen that lists the steps
 of the installation.
  Verify Installed Packages
  Network Bridge Configuration
 The Verify Installed Packages installs the following:
 libvirt
 libvirt-python
 vm-install
 kvm
 virt-manager
 virt-viewer
 
 The packages install without error. The process never asks to setup a
 Network Bridge Configuration. It then tells me
 For installing KVM guests, reboot the machine to load the necessary
 drivers., and then I am back to the main yast menu.

With SLES12, that message is now incorrect.  For SLES11 SP3 it is necessary to 
reboot for things to work.  You won't be asked to configure a bridge at this 
point, and it's not needed, and won't work.

 Could you point me to documentation on a macvtap setup?

The easiest way is just to use virt-manager to create a new KVM guest.  It 
should automatically create the macvtap interface for you.


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM network setup on SUSE SLES

[Mark Post]

 On 1/6/2015 at 05:03 PM, Brad Gartside brad.garts...@gmail.com wrote: 
 I am setting up a KVM proof of concept system on a z/Linux guest running on
 z/VM. I have tried using the GUI and text versions of yast to Install
 Hypervisor and Tools, and the install process always skips over the
 Network Bridge Configuration setup step.

Which install process are you talking about here?

 The behavior has been repeated
 on SLES 11 SP3 GA and SLES 12 GA. The z/Linux KVM host guest has a Layer 2
 OSA networking defined.
 
 Has anyone else run into network problems setting up KVM on SLES?

Not a problem, exactly, but you need to be aware that using macvtap is the only 
thing that is going to work for you, due to the nature of the OSA hardware 
itself.  The virt-manager interface should set that up for you automatically, 
however.


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM on IBM System z

[Mark Post]

 On 10/17/2012 at 01:03 AM, Tobias Doerkes tdoer...@hotmail.com wrote: 
 Hi all,
 one more question regarding my KVM tests:
 
 Does anybody know wether i can specifiy the virtio_net driver and interface 
 for a kvm machine in SLES parmfile to automate the startup of the installer 
 image? 

I would be very surprised if the startup of the installer would have any idea 
how to handle anything other than what gets listed in the prompt:
1) OSA-2 or OSA Express
2) Hipersockets
3) Channel To Channel (CTC)
4) ESCON
5) Inter-User Communication Vehicle (IUCV)


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM on IBM System z

[Tobias Doerkes]

Hi all,
one more question regarding my KVM tests:

Does anybody know wether i can specifiy the virtio_net driver and interface for 
a kvm machine in SLES parmfile to automate the startup of the installer image? 

Manual definition and dhcp networking via http are working without any problem. 
But i want to configure the installer image by passing the instserver 
information via parmfile. Problem is, that during processing of the parmfile a 
network device is required. But using manual=1 skips this requirement.

See massages below:

Starting hardware detection... ok
(If a driver is not working for you, try booting with 
brokenmodules=driver_name.)

Virtio Storage 0
  drivers: virtio_blk*
Virtio Ethernet Card 0
  drivers: virtio_net*

Please select the type of your network device.

1) OSA-2 or OSA Express
2) Hipersockets

3) Channel To Channel (CTC)
4) ESCON
5) Inter-User Communication Vehicle (IUCV)



*** Could not find the SUSE Linux Enterprise Server 11 Repository.

Activating manual setup program.

 Linuxrc v3.3.81 (Kernel 3.0.13-0.27-default) 

Main Menu

1) Start Installation
2) Settings
3) Expert
4) Exit or Reboot

 1

Start Installation

1) Start Installation or Update
2) Boot Installed System
3) Start Rescue System

 1

Choose the source medium.

1) DVD / CD-ROM
2) Network

 2

Choose the network protocol.

1) FTP
2) HTTP
3) NFS
4) SMB / CIFS (Windows Share)
5) TFTP

 2
Detecting and loading network drivers

Automatic configuration via DHCP?

1) Yes
2) No

 1
Sending DHCP request...

Enter the IP address of the HTTP server
[x.x.x.x]

Enter the directory on the server
[instserver/sles-11-sp2-s390x-iso/]

Do you need a username and password to access the HTTP server?

1) Yes
2) No

 2

Use a HTTP proxy?

1) Yes
2) No

 2
Loading Installation System (1/6) -  100%
squashfs: version 4.0 (2009/01/31) Phillip Lougher
Loading Installation System (2/6) -  100%
Loading Installation System (3/6) -  100%
Loading Installation System (4/6) -  100%
Loading Installation System (5/6) -  100%
Loading Installation System (6/6) -  100%
Reading Driver Update...

No new Driver Updates found

Select the display type.

1) X11
2) VNC
3) SSH
4) ASCII Console


Kind regards,

Tobias
  
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Antwort: Re: KVM on IBM System z

[Carsten Otte]

Your qemu process should have kvm file descriptors open in /proc/pidfds/
and you should see a debug area in /sys/kernel/debug/s390dbf/kvm-pid/ 

with kind regards
Carsten Otte
System z firmware development / Boeblingen lab
---
Every revolution was first a thought in one man's mind; 
and when the same thought occurs to another man, it is the key to that era. 
 - Ralph Waldo Emerson, Essays: First Series, 1841

-Linux on 390 Port LINUX-390@vm.marist.edu schrieb: -
An: LINUX-390@vm.marist.edu
Von: Tobias Doerkes 
Gesendet von: Linux on 390 Port 
Datum: 11.10.2012 07:15
Betreff: Re: KVM on IBM System z

Hi all,

one more question regarding KVM on IBM System z:
Is there a way to check wether KVM is using hardware virtualisation (SIE 
instruction)?

I installed SLES 11 and virt-host-validate is missing. In FC 17 it returns only 
software virtualisation:

  QEMU: Checking for hardware virtualization : 
WARN (Only emulated 
CPUs are available, performance will be significantly limited)
  QEMU: Checking for device /dev/vhost-net   : 
PASS
  QEMU: Checking for device /dev/net/tun : 
PASS
   LXC: Checking for Linux = 2.6.26 : 
PASS

But i think virt-host-validate in FC17 has no support for s390x. So i want to 
check wether SIE is used or not.

Kind regards,

Tobias.

      
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM on IBM System z

[Tobias Doerkes]

Hi all,

many thanks to Carsten. And i think i found one more way:

[root@linux qemu]# ./bin/qemu-system-s390x -monitor stdio -machine 
type=s390-virtio,accel=kvm
VNC server running on `127.0.0.1:5900'
QEMU 1.2.0 monitor - type 'help' for more information
(qemu) info kvm
kvm support: enabled
(qemu)

By the way, is there any support for the virtual nics in the current s390x 
distros? I suppose qeth and iucv will not work with the network interfaces.

Kind regards,

Tobias

 Your qemu process should have kvm file descriptors open in /proc/pidfds/
 and you should see a debug area in /sys/kernel/debug/s390dbf/kvm-pid/ 
 
 with kind regards
 Carsten Otte
 System z firmware development / Boeblingen lab
 ---
 Every revolution was first a thought in one man's mind; 
 and when the same thought occurs to another man, it is the key to that era. 
  - Ralph Waldo Emerson, Essays: First Series, 1841
 

  
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM on IBM System z

[Tobias Doerkes]

Hi all,

one more question regarding KVM on IBM System z:
Is there a way to check wether KVM is using hardware virtualisation (SIE 
instruction)?

I installed SLES 11 and virt-host-validate is missing. In FC 17 it returns only 
software virtualisation:

  QEMU: Checking for hardware virtualization : 
WARN (Only emulated 
CPUs are available, performance will be significantly limited)
  QEMU: Checking for device /dev/vhost-net   : 
PASS
  QEMU: Checking for device /dev/net/tun : 
PASS
   LXC: Checking for Linux = 2.6.26 : 
PASS

But i think virt-host-validate in FC17 has no support for s390x. So i want to 
check wether SIE is used or not.

Kind regards,

Tobias.

  
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM on IBM System z

[David K. Kelly]

inline: graycol.gifinline: pic32170.gifinline: ecblank.gif

Re: KVM on IBM System z

[Mark Post]

 On 9/26/2012 at 11:37 AM, Tobias Doerkes tdoer...@hotmail.com wrote: 

 for several years there is s390x support in qemu and the fedora guys have 
 qemu-kvm on their distro. 
 Has anybody testet KVM on IBM System z yet?

I've played with it a little bit.  The networking setup can be a bit touchy, 
but that might just be my inexperience with KVM in general.  There's 
documentation spread all over the Internet, but not much of it has anything to 
say about doing it in a mainframe environment.  I'm hoping to fix some of that 
in the coming months.  If anyone has anything they can share to speed that up, 
let me know.  :)


Mark Post

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM on IBM System z

[Philipp Kern]

On Wed, Sep 26, 2012 at 03:37:45PM +, Tobias Doerkes wrote:
 for several years there is s390x support in qemu and the fedora guys have
 qemu-kvm on their distro.
 Has anybody testet KVM on IBM System z yet?

Works as advertised with kernel/initrd boot. It does not have mainfraime
I/O, hence you get normal virtio disks and NICs. SUSE has patches to
zipl to boot from virtio disks. On the Debian side we're still thinking
whether we should apply them or not, given that they won't be integrated
by IBM. (A silly upstream it is.)

Kind regards
Philipp Kern

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/

Re: KVM

[Alan Altmark]

On Wednesday, 12/03/2008 at 02:01 EST, Shawn Wells [EMAIL PROTECTED]
wrote:
 Come on now, haven't you heard the Mainframe is dead? ;)

 While I certainly can't speak for entire companies, I'm aware not aware
 of any such efforts.

I can represent IBM with respect to z/VM:  IBM is not getting rid of z/VM
in favor of KVM.

IBM is fully aware of KVM and it's potential intersection with z/VM should
a System z version ever emerge.  It does not in any way devalue z/VM.

Anyone who says otherwise is ill-informed or simply trying to create FUD.
If you hear such a message from an IBMer, please send me details of the
contact and I will see to it that they receive a proper education.

Alan Altmark
z/VM Development
IBM Endicott

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Re: KVM

[Carsten Otte]

Am Thu, 4 Dec 2008 09:59:30 -0500
schrieb Alan Altmark [EMAIL PROTECTED]:
 I can represent IBM with respect to z/VM:  IBM is not getting rid of z/VM
 in favor of KVM.

 IBM is fully aware of KVM and it's potential intersection with z/VM should
 a System z version ever emerge.  It does not in any way devalue z/VM.

 Anyone who says otherwise is ill-informed or simply trying to create FUD.
 If you hear such a message from an IBMer, please send me details of the
 contact and I will see to it that they receive a proper education.
I second that. For those interrested in the current status
of KVM on the platform, see
http://kvm.qumranet.com/kvmwiki/KvmForum2008?action=AttachFiledo=gettarget=kdf2008_17.pdf

This presentation is from spring 2008, the stop machine run
and memory overcommitment issues have been fixed by now as
of 2.6.28-rc.

If you want to try it out, see:
http://www.ibm.com/developerworks/linux/linux390/kuli.html

so long,
Carsten

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390