cloudstack-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Wido den Hollander <w...@widodh.nl>
Subject Re: Hyper-V clock on KVM node - CLOUDSTACK-8978
Date Thu, 29 Oct 2015 08:48:21 GMT


On 27-10-15 23:42, Josh Harshman wrote:
> I noticed around line 1949 in LibvirtComputingResource.java, there is some commented
out code for enabling hyperV enlightenment.  Was there an issue with enabling this behavior?
> 

I don't know why this was disabled. Maybe somebody else knows, but I
can't tell for sure.

Wido

> Code Snip:
> //for rhel 6.5 and above, hyperv enlightment feature is added
> /*
>   * if (vmTO.getOs().contains("Windows Server 2008") && hostOsVersion != null
&& ((hostOsVersion.first() == 6 &&     hostOsVersion.second()       >=
5) || (hostOsVersion.first() >= 7))) {
>    *    LibvirtVMDef.HyperVEnlightenmentFeatureDef hyv = new LibvirtVMDef.HyperVEnlightenmentFeatureDef();
>    *    hyv.setRelaxed(true);
>    *    features.addHyperVFeature(hyv);
>    * }
>    */
> 
> 
> 
> ________________________________________
> From: Wido den Hollander <wido@widodh.nl>
> Sent: Monday, October 26, 2015 9:05 AM
> To: dev@cloudstack.apache.org
> Subject: Re: Hyper-V clock on KVM node - CLOUDSTACK-8978
> 
> 
> 
> Josh Harshman
> Cloud Engineer
> Security+
> Intrinium
> 
> On 24-10-15 02:15, Josh Harshman wrote:
>> Just looking for some feedback on this idea for an improvement.
>>
>>
>> Currently, for any windows guest OS, we set RTC for the clock source. This seems
to contribute to clock-drift issues leading to BSODs on Windows Server 2008 R2 Guests.
>>
>> I believe a possible improvement would be to set Windows guests to use the hyper-v
clock. To do this we will need CloudStack to generate the appropriate Libvirt XML which would
look somewhat like the following:
>>
>> <features>
>> <hyperv>
>> <relaxed state='on'/>
>> <vapic state='on'/>
>> <spinlocks state='on' retries='8191'/>
>> </hyperv>
>> <features/>
>> <clock ...>
>> <timer name='hypervclock' present='yes'/>
>> </clock>
>>
>>
>> I believe the following code snippet from LibvirtComputingResource.java could be
a starting place to implement this.
>>
>>
>> if (vmTO.getOs().startsWith("Windows")) {
>> clock.setClockOffset(ClockDef.ClockOffset.LOCALTIME);
>> clock.setTimer("rtc", "catchup", null);
>>
>>
>> Let me know what you all think about this change.
>> Is there a better place / way to implement this?
>>
> 
> That would indeed be the place to fix this. It's currently rather
> hardcoded, but that would be the place to go.
> 
> Wido
> 
>>
>>
>>
>>
>> Josh Harshman
>>
>> Cloud Engineer
>> Security+
>>
>>
>> Intrinium
>> Tel: (509) 465-1234 x5259
>> Fax: (866) 565-4578
>> Lync / Skype: Josh.Harshman@intrinium.com
>> Web: http://intrinium.com<http://intrinium.com/>
>>
>>
>> [Intrinium Long Sig Logo]<http://intrinium.com/>
>>
>>
>> [Facebook]<http://www.facebook.com/pages/intrinium/154891544582822> [Twitter]
<http://twitter.com/intrinium>  [Linkedin] <http://www.linkedin.com/company/intrinium_networks_it_security?trk=fc_badge>
 [Youtube] <http://www.youtube.com/user/intrinium>  [Blog] <http://intrinium.com/blog/>
>>
>> Information Security and Compliance Consulting | Managed IT and Security Services
| Cloud Services
>>
>>
>>
>>
>>
>> ________________________________
>>
>> This email and any files transmitted with it are confidential and intended solely
for the use of the individual or entity to whom they are addressed. If you have received this
email in error please notify the system manager. This message contains confidential information
and is intended only for the individual named. If you are not the named addressee you should
not disseminate, distribute or copy this e-mail. Please notify the sender immediately by e-mail
if you have received this e-mail by mistake and delete this e-mail from your system. If you
are not the intended recipient you are notified that disclosing, copying, distributing or
taking any action in reliance on the contents of this information is strictly prohibited.
>>

Mime
View raw message