Test Xen 4.8 RC5 instable 04.11.16

Test Xen 4.8 RC5 instable 04.11.16

[Juergen Schinker]

* Hardware:
 Intel(R) Xeon(R) CPU E3-1220L V2 @ 2.30GHz
 Sandisk SSD 1T
 32G Ram
* Software:

Debian Stretch/testing is dom0
 
* Guest operating systems:

Guests Ubuntu 14.04 LTS Debian Stretch/Sid 

Ubuntu 16.10 yakity yak with latest 4.8 kernel works 
 
* Functionality tested:
xl 
creating booting 
pygrub
credit2

* Comments:

 I compiled the latest vanilla kernel for gpu-passthrough with single gpu -

rc5 runs ca. 1h then freeze - which log should i check for ?
 

host                   : xen
release                : 4.8.5-xen
version                : #1 SMP PREEMPT Mon Oct 31 20:31:38 GMT 2016
machine                : x86_64
nr_cpus                : 4
max_cpu_id             : 3
nr_nodes               : 1
cores_per_socket       : 2
threads_per_core       : 2
cpu_mhz                : 2312
hw_caps                : b7ebfbff:77bae3ff:28100800:00000001:00000001:00000281:00000000:00000100
virt_caps              : hvm hvm_directio
total_memory           : 32711
free_memory            : 4133
sharing_freed_memory   : 0
sharing_used_memory    : 0
outstanding_claims     : 0
free_cpus              : 0
xen_major              : 4
xen_minor              : 8
xen_extra              : .0-rc
xen_version            : 4.8.0-rc
xen_caps               : xen-3.0-x86_64 xen-3.0-x86_32p hvm-3.0-x86_32 hvm-3.0-x86_32p hvm-3.0-x86_64 
xen_scheduler          : credit2
xen_pagesize           : 4096
platform_params        : virt_start=0xffff800000000000
xen_changeset          : Mon Oct 31 13:21:56 2016 +0000 git:496673a
xen_commandline        : placeholder sched=credit2 iommu=1
cc_compiler            : gcc (Debian 6.2.0-9) 6.2.0 20161019
cc_compile_by          : root
cc_compile_domain      : 
cc_compile_date        : Thu Nov  3 20:55:16 GMT 2016
build_id               : de5b411bee8c8590609ddfe7f425344deee707ba
xend_config_format     : 4

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: Test Xen 4.8 RC5 instable 04.11.16

[Wei Liu]

On Fri, Nov 04, 2016 at 11:13:29AM +0000, Juergen Schinker wrote:
> * Hardware:
>  Intel(R) Xeon(R) CPU E3-1220L V2 @ 2.30GHz
>  Sandisk SSD 1T
>  32G Ram
> * Software:
> 
> Debian Stretch/testing is dom0
>  
> * Guest operating systems:
> 
> Guests Ubuntu 14.04 LTS Debian Stretch/Sid 
> 
> Ubuntu 16.10 yakity yak with latest 4.8 kernel works 
>  
> * Functionality tested:
> xl 
> creating booting 
> pygrub
> credit2
> 
> * Comments:
> 
>  I compiled the latest vanilla kernel for gpu-passthrough with single gpu -
> 
> rc5 runs ca. 1h then freeze - which log should i check for ?
>  

If the guest freezes, you can have a look at guest serial output, Dom0
dmesg and xen dmesg (xl dmesg).

If the host freezes, you will need to setup a serial console and uses
another machine to capture the log as it goes.

> 
> host                   : xen
> release                : 4.8.5-xen
> version                : #1 SMP PREEMPT Mon Oct 31 20:31:38 GMT 2016
> machine                : x86_64
> nr_cpus                : 4
> max_cpu_id             : 3
> nr_nodes               : 1
> cores_per_socket       : 2
> threads_per_core       : 2
> cpu_mhz                : 2312
> hw_caps                : b7ebfbff:77bae3ff:28100800:00000001:00000001:00000281:00000000:00000100
> virt_caps              : hvm hvm_directio
> total_memory           : 32711
> free_memory            : 4133
> sharing_freed_memory   : 0
> sharing_used_memory    : 0
> outstanding_claims     : 0
> free_cpus              : 0
> xen_major              : 4
> xen_minor              : 8
> xen_extra              : .0-rc
> xen_version            : 4.8.0-rc
> xen_caps               : xen-3.0-x86_64 xen-3.0-x86_32p hvm-3.0-x86_32 hvm-3.0-x86_32p hvm-3.0-x86_64 
> xen_scheduler          : credit2
> xen_pagesize           : 4096
> platform_params        : virt_start=0xffff800000000000
> xen_changeset          : Mon Oct 31 13:21:56 2016 +0000 git:496673a
> xen_commandline        : placeholder sched=credit2 iommu=1
> cc_compiler            : gcc (Debian 6.2.0-9) 6.2.0 20161019
> cc_compile_by          : root
> cc_compile_domain      : 
> cc_compile_date        : Thu Nov  3 20:55:16 GMT 2016
> build_id               : de5b411bee8c8590609ddfe7f425344deee707ba
> xend_config_format     : 4

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: Test Xen 4.8 RC5 instable 04.11.16

[Dario Faggioli]

[-- Attachment #1.1: Type: text/plain, Size: 1209 bytes --]

On Fri, 2016-11-04 at 11:18 +0000, Wei Liu wrote:
> On Fri, Nov 04, 2016 at 11:13:29AM +0000, Juergen Schinker wrote:
> > 
> > * Functionality tested:
> > xl 
> > creating booting 
> > pygrub
> > credit2
> > 
> > * Comments:
> > 
> >  I compiled the latest vanilla kernel for gpu-passthrough with
> > single gpu -
> > 
> > rc5 runs ca. 1h then freeze - which log should i check for ?
> >  
> 
> If the guest freezes, you can have a look at guest serial output,
> Dom0
> dmesg and xen dmesg (xl dmesg).
> 
> If the host freezes, you will need to setup a serial console and uses
> another machine to capture the log as it goes.
> 
Also of interest (and in the meanwhile): have you done the same in your
previous testing of previous rc-s and they did work?

In other word, would you say it is something in rc5 that is introducing
this behavior, which was not happening of earlier versions?

Thanks and Regards,
Dario
-- 
<<This happens because I choose it to happen!>> (Raistlin Majere)
-----------------------------------------------------------------
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Senior Software Engineer, Citrix Systems R&D Ltd., Cambridge (UK)

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

[-- Attachment #2: Type: text/plain, Size: 127 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: Test Xen 4.8 RC5 instable 04.11.16

[Juergen Schinker]

----- On 4 Nov, 2016, at 17:29, Dario Faggioli dario.faggioli@citrix.com wrote:

> Also of interest (and in the meanwhile): have you done the same in your
> previous testing of previous rc-s and they did work?
> 
Yes and no it didn't work

> In other word, would you say it is something in rc5 that is introducing
> this behavior, which was not happening of earlier versions?
 
it was introduced in rc4 and I consider rc3 stable.

the rtds sched has not been tested.

J

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: Test Xen 4.8 RC5 instable 04.11.16

[Dario Faggioli]

[-- Attachment #1.1: Type: text/plain, Size: 1304 bytes --]

On Fri, 2016-11-04 at 18:11 +0000, Juergen Schinker wrote:
> ----- On 4 Nov, 2016, at 17:29, Dario Faggioli dario.faggioli@citrix.
> com wrote:
> > Also of interest (and in the meanwhile): have you done the same in
> > your
> > previous testing of previous rc-s and they did work?
> > 
> Yes and no it didn't work
> 
Ok.

> > In other word, would you say it is something in rc5 that is
> > introducing
> > this behavior, which was not happening of earlier versions?
>  
> it was introduced in rc4 and I consider rc3 stable.
> 
Ok. Well, we absolutely need serial output (either of the guest, of the
host, or of both) to be able to tell.

Info about how to setup them here:

https://wiki.xenproject.org/wiki/Xen_Serial_Console
https://wiki.xenproject.org/wiki/Xen_FAQ_Console

> the rtds sched has not been tested.
> 
I never asked about RTDS. I am interested in and did ask why you tested
Credit2, how you tested it, and how you do find it, though, if you're
keen on sharing that. :-D

Thanks and Regards,
Dario
-- 
<<This happens because I choose it to happen!>> (Raistlin Majere)
-----------------------------------------------------------------
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Senior Software Engineer, Citrix Systems R&D Ltd., Cambridge (UK)

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

[-- Attachment #2: Type: text/plain, Size: 127 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

Re: Test Xen 4.8 RC5 instable 04.11.16

[Juergen Schinker]

----- On 4 Nov, 2016, at 18:31, Dario Faggioli dario.faggioli@citrix.com wrote:

> On Fri, 2016-11-04 at 18:11 +0000, Juergen Schinker wrote:
>> ----- On 4 Nov, 2016, at 17:29, Dario Faggioli dario.faggioli@citrix.
>> com wrote:
>> > Also of interest (and in the meanwhile): have you done the same in
>> > your
>> > previous testing of previous rc-s and they did work?
>> > 
>> Yes and no it didn't work
>> 
> Ok.
> 
>> > In other word, would you say it is something in rc5 that is
>> > introducing
>> > this behavior, which was not happening of earlier versions?
>>  
>> it was introduced in rc4 and I consider rc3 stable.
>> 
> Ok. Well, we absolutely need serial output (either of the guest, of the
> host, or of both) to be able to tell.
> 
> Info about how to setup them here:
> 
> https://wiki.xenproject.org/wiki/Xen_Serial_Console
> https://wiki.xenproject.org/wiki/Xen_FAQ_Console
> 
>> the rtds sched has not been tested.
>> 
> I never asked about RTDS. I am interested in and did ask why you tested
> Credit2, how you tested it, and how you do find it, though, if you're
> keen on sharing that. :-D

I consider credit2 as the new default and tested it while using it with a lot of guests and different load; I also played

with a lot of vcpu-set and vcpu-pin etc.

credit2 is fantastic and all guests work well balanced and snappy...

J

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel