incubator-cloudstack-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mice Xia" <mice_...@tcloudcomputing.com>
Subject RE: [ACS41] VM Snapshot and possibility to introduce it in 4.1
Date Wed, 19 Dec 2012 06:17:29 GMT
[David]
Couple of questions around memory - as you note XenServer free doesn't have support for memory
snapshots - does that increase the number of code paths for this feature?
[/David]

Yes, but not much, just call different XAPIs depending on if it's a memory snapshot, either
vm.snapshotAsync or vm.checkpointAsync.

[David]
Also - I imagine snapshotting memory is dramatically more intensive from a diskio perspective.
(so a machine with 64GB of RAM now has a 64GB write job) - many moons ago, we specifically
modified qemu to not do that because of the performance impact. Any idea what the impact is?
[/David]

Thanks for the reminder, I haven’t tested the performance, simple functional tests are based
on 512M/2G memory VM. It's no good if it will impact other guest VMs on the same host.

-Mice

-----Original Message-----
From: David Nalley [mailto:david@gnsa.us] 
Sent: Saturday, December 15, 2012 1:19 AM
To: cloudstack-dev@incubator.apache.org
Subject: Re: [ACS41] VM Snapshot and possibility to introduce it in 4.1

On Fri, Dec 14, 2012 at 7:27 AM, Mice Xia <mice_xia@tcloudcomputing.com> wrote:
> Hi, folks,
>
> I’d like to discuss the possibility of introducing feature VM Snapshot in CS version
4.1 and I need some inputs/ideas for the rest of work.
>
> The background/spec/design is in https://cwiki.apache.org/confluence/display/CLOUDSTACK/VM+Snapshots.
> The implementation is very straightforward (with KISS principle) , codes are in branch
vm-snapshot, based on 4.0.0.
>

Wow the spec looks good.

Couple of questions around memory - as you note XenServer free doesn't have support for memory
snapshots - does that increase the number of code paths for this feature?

Also - I imagine snapshotting memory is dramatically more intensive from a diskio perspective.
(so a machine with 64GB of RAM now has a 64GB write job) - many moons ago, we specifically
modified qemu to not do that because of the performance impact. Any idea what the impact is?

--David
Mime
View raw message