> On 2012年11月19日 17:30, Osier Yang wrote:
> > Hi,
> >
> > This proposal is trying to figure out a solution for migration
> > of domain which uses LUN behind vHBA as disk device (QEMU
> > emulated disk only at this stage). And other related NPIV
> > improvements which are not related with migration.
No semantic change, but prepare for a new mode of parsing where
a new _BRANCH flag requests that the parse look up the existing
snapshot to branch from.
* src/conf/snapshot_conf.h (VIR_DOMAIN_SNAPSHOT_PARSE_BRANCH): New
flag, unused for now.
(virDomainSnapshotDefParseString): Add parameter.
* src/
When asked to parse a branch snapshot XML definition, we have to
piece together the definition of the new snapshot from parts of
the branch point, as well as run some sanity checks that the
branches are compatible. This patch is rather restrictive in
what it allows; depending on effort and future
Upcoming patches for revert-and-clone branching of snapshots need
to be able to copy a domain definition; make this step reusable.
* src/conf/domain_conf.h (virDomainDefCopy): New prototype.
* src/conf/domain_conf.c (virDomainObjCopyPersistentDef): Split...
(virDomainDefCopy): ...into new function
Right now, libvirt refuses to revert to the state at the time
of an external snapshot, because doing so would reset things so
that the next time the domain boots, we are using the backing
file; but modifying the backing file invalidates all qcow2
files that are based on top of it. There are three
v1 was here:
https://www.redhat.com/archives/libvir-list/2012-November/msg00484.html
Since then, I've folded in review comments on 1; 2-4 are new.
Documentation is still pending (split out of 1 based on the v1
review). Also, I'm still writing and testing the actual qemu use
of the new snapshot_co
On 11/19/2012 09:06 AM, Peter Krempa wrote:
> This patch adds support for reverting of external snapshots. The support
> is somewhat limited yet (you can only revert to a snapshot that has no
> children or delete the children that would have their image chains
> invalidated).
>
> While reverting a
On 11/19/2012 09:06 AM, Peter Krempa wrote:
> Some hypervisors require a respawn of the hypervisor to allow reverting
> to some snapshot states. This patch adds flag to remove the default
> safe approach to not allow this. When this flag is specified the
> hypervisor driver should re-emit events to
On 11/19/2012 09:06 AM, Peter Krempa wrote:
> The current snapshot reverting api supported changing the state of the
> machine after the snapshot was reverted to either started or paused.
>
> This patch adds the ability to revert the state but to stopped state.
> ---
> Fixed comment regarding use
On 11/19/2012 09:06 AM, Peter Krempa wrote:
> I'm reposting another version of the series. This version fixes small mistakes
> in patches 1-3 and reposts patch 4 as it got missing due to a sending error.
>
>
> Peter Krempa (4):
> qemu: Split out guts of qemuDomainSaveImageStartVM() to allow reu
On 11/19/2012 06:48 AM, Peter Krempa wrote:
> On 11/15/12 00:36, Eric Blake wrote:
>> Now that we can filter on this information, we should also make
>> it easy to get at.
>>
>> * tools/virsh-snapshot.c (cmdSnapshotInfo): Add another output
>> row, and switch to XPath queries rather than strstr.
>
On 11/19/2012 03:51 PM, Laine Stump wrote:
On 11/19/2012 03:27 PM, Gene Czarcinski wrote:
On 11/19/2012 01:06 PM, Laine Stump wrote:
Here is an open BZ that mentions the idea of adding similar
functionality for dnsmasq (see comment 8, in particular):
https://bugzilla.redhat.com/show_bug.cg
On 11/19/2012 04:26 AM, Peter Krempa wrote:
> On 11/15/12 00:36, Eric Blake wrote:
>> Snapshot filtering based on types is useful enough to add
>> back-compat support into virsh. It is also rather easy - all
>> versions of libvirt that don't understand the new filter flags
>> already gave us suffi
On 11/19/2012 02:24 PM, Laine Stump wrote:
1. In a manner similar to what is done for IPV6, add ip6tables rules
>>>to permit virtual systems to communicate via a defined virtual
>>>interface which has no gateway addresses defined. This does mean that
>>>virtual systems will not be able to commu
On 11/19/2012 03:27 PM, Gene Czarcinski wrote:
> On 11/19/2012 01:06 PM, Laine Stump wrote:
>> Here is an open BZ that mentions the idea of adding similar
>> functionality for dnsmasq (see comment 8, in particular):
>>
>>https://bugzilla.redhat.com/show_bug.cgi?id=824573
> I also found Comment
On 11/19/2012 01:06 PM, Laine Stump wrote:
Here is an open BZ that mentions the idea of adding similar
functionality for dnsmasq (see comment 8, in particular):
https://bugzilla.redhat.com/show_bug.cgi?id=824573
I also found Comment 2 where Dave Allan says "The intent is that people
who want
On 11/19/2012 01:50 PM, Gene Czarcinski wrote:
> SInce I wrote this, I have "Plan B" about a half to two-thirds
> complete as far as the coding goes. I have this "feeling" that
> postponing the conf-file changes might be better.
>
> On 11/19/2012 01:06 PM, Laine Stump wrote:
>> On 11/19/2012 08:31
SInce I wrote this, I have "Plan B" about a half to two-thirds complete
as far as the coding goes. I have this "feeling" that postponing the
conf-file changes might be better.
On 11/19/2012 01:06 PM, Laine Stump wrote:
On 11/19/2012 08:31 AM, Gene Czarcinski wrote:
Plan A:
This consists of
On 11/19/2012 08:31 AM, Gene Czarcinski wrote:
> Plan A:
>
> This consists of the set of patches I have submitted for conf-file,
> DHCPv6, etc. The response has been a resounding SILENCE. Thus, I am
> assuming that there is some reluctance to adopting these changes in
> their current form. This
On 11/19/12 18:19, Michal Privoznik wrote:
On 19.11.2012 17:06, Peter Krempa wrote:
I'm reposting another version of the series. This version fixes small mistakes
in patches 1-3 and reposts patch 4 as it got missing due to a sending error.
Peter Krempa (4):
qemu: Split out guts of qemuDomai
On 19.11.2012 17:06, Peter Krempa wrote:
> I'm reposting another version of the series. This version fixes small mistakes
> in patches 1-3 and reposts patch 4 as it got missing due to a sending error.
>
>
> Peter Krempa (4):
> qemu: Split out guts of qemuDomainSaveImageStartVM() to allow reuse
This approach implemented in previous patches is not trivial
and deserves small description.
---
src/util/virnetdevbandwidth.c | 68 +---
1 files changed, 62 insertions(+), 6 deletions(-)
diff --git a/src/util/virnetdevbandwidth.c b/src/util/virnetdevbandwidt
Status file keeps track of class_id and floor_sum. It's better
to keep it updated in case libvirtd is killed.
---
src/network/bridge_driver.c | 14 ++
1 files changed, 14 insertions(+), 0 deletions(-)
diff --git a/src/network/bridge_driver.c b/src/network/bridge_driver.c
index 8dc9d
Currently, we are only keeping a inactive XML configuration
in status dir. This is no longer enough as we need to keep
this class_id attribute so we don't overwrite old entries
when the daemon restarts. However, since there has already
been a libvirt release which has just as root element,
and we
SFQ is qdisc which doesn't really shape any traffic but
'just' re-arrange packets in sending buffer so no stream
starve. The goal is to ensure fairness. There is basically
only one configuration parameter (perturb) which is set to
advised value of 10.
---
src/util/virnetdevbandwidth.c | 10
With current implementation, class ID is just incremented. This can
lead to its exhaustion as tc accepts only 16 bits long identifiers.
Hence, it's better if we allow class ID to be reused. To keep track
which IDs are free and which are taken virBitmap is used. This requires
network status file to
This will be used whenever a NIC with guaranteed throughput is to
be plugged into a bridge. It will adjust the average throughput of
non guaranteed NICs (classid 1:2) to meet new requirements.
---
src/util/virnetdevbandwidth.c | 49 +
src/util/virnetdevban
Interfaces keeps a class_id, which is an ID from which bridge
part of QoS settings is derived. We need to store class_id
in domain status file, so we can later pass it to
virNetDevBandwidthUnplug.
---
src/conf/domain_conf.c |4 +-
src/qemu/qemu_domain.c | 66 +
Network should be notified if we plug in or unplug an
interface, so it can perform some action, e.g. set/unset
network part of QoS.
---
src/Makefile.am |7 ++-
src/conf/domain_conf.h |1 +
src/conf/network_conf.c |1 +
src/conf/network_conf.h |3 +
src/libv
This is however supported only on domain interfaces with
type='network'. Moreover, target network needs to have at least
inbound QoS set. This is required by hierarchical traffic shaping.
>From now on, the required attribute for is either 'average'
(old) or 'floor' (new). This new attribute can b
These classes can borrow unused bandwidth. Basically,
only egress qdsics can have classes, therefore we can
do this kind of traffic shaping only on host's outgoing,
that is domain's incoming traffic.
---
src/lxc/lxc_process.c |3 ++-
src/network/bridge_driver.c |3 ++-
src/qemu/q
These set bridge part of QoS when bringing domain's interface up.
Long story short, if there's a 'floor' set, a new QoS class is created.
ClassID MUST be unique within the bridge and should be kept for
unplug phase.
---
po/POTFILES.in|1 +
src/util/virnetdevbandwidth.c | 178 +
This is a rework of my one year old patchset [1].
If you are brave enough and start reviewing it, you may want to read
the last patch first. The whole magic hidden behind the scene is
documented there. For now, the only supported combination where this
works is virtual interface connected to virtu
This patch adds support for reverting of external snapshots. The support
is somewhat limited yet (you can only revert to a snapshot that has no
children or delete the children that would have their image chains
invalidated).
While reverting an external snapshot, the domain has to be taken offline
The current snapshot reverting api supported changing the state of the
machine after the snapshot was reverted to either started or paused.
This patch adds the ability to revert the state but to stopped state.
---
Fixed comment regarding use of this flag on transient domains.
---
include/libvirt/
The workhorse part of qemuDomainSaveImageStartVM can be reused while
loading external snapshots. This patch splits the code out into a new
function qemuDomainSaveImageLoad that is free of setting lifecycle
events.
---
No changes, ACKed.
---
src/qemu/qemu_driver.c | 53 +
I'm reposting another version of the series. This version fixes small mistakes
in patches 1-3 and reposts patch 4 as it got missing due to a sending error.
Peter Krempa (4):
qemu: Split out guts of qemuDomainSaveImageStartVM() to allow reuse
snapshot: Add flag VIR_DOMAIN_SNAPSHOT_REVERT_STOPP
Some hypervisors require a respawn of the hypervisor to allow reverting
to some snapshot states. This patch adds flag to remove the default
safe approach to not allow this. When this flag is specified the
hypervisor driver should re-emit events to allow management apps to
reconnect.
This flag is m
Hello all,
Someone can explain me how can I generate connection with credential like
user and pass (ex: root\passw0rd)
I use Java API, and use this following rows for connection with local
host :
conn = new Connect("qemu:///system", false);
Thanks in advance
On 11/11/2012 10:25 AM, Toralf Förster wrote:
> I've a stable Gentoo Linux, the build log is attached.
>
It's hard to say from the log, but you can check yourselves what the
problem is by running the particular test like this:
VIR_TEST_DEBUG=1 ./virdrivermoduletest
in the tests/ directory after
On 11/15/12 00:36, Eric Blake wrote:
Now that we can filter on this information, we should also make
it easy to get at.
* tools/virsh-snapshot.c (cmdSnapshotInfo): Add another output
row, and switch to XPath queries rather than strstr.
---
tools/virsh-snapshot.c | 57 ++
Plan A:
This consists of the set of patches I have submitted for conf-file,
DHCPv6, etc. The response has been a resounding SILENCE. Thus, I am
assuming that there is some reluctance to adopting these changes in
their current form. This especially includes the conf-file changes.
While I b
On 11/15/12 00:36, Eric Blake wrote:
Relatively straight-forward. And since qemu was already using
VIR_DOMAIN_SNAPSHOT_FILTERS_ALL, with 6 different APIs all calling
into this common code, I've instantly added all 5 flags to 6 APIs.
* src/conf/snapshot_conf.h (VIR_DOMAIN_SNAPSHOT_FILTERS_ALL):
On 11/15/12 00:36, Eric Blake wrote:
Snapshot filtering based on types is useful enough to add
back-compat support into virsh. It is also rather easy - all
versions of libvirt that don't understand the new filter flags
already gave us sufficient information in a single XML field
to reconstruct a
On 11/15/12 00:20, Eric Blake wrote:
On 11/14/2012 11:43 AM, Eric Blake wrote:
Bleah. Raw XML parsing. Wouldn't it be easier in and cleaner convert
this piece code to use the XML parser and xpath?
Not the first time we've done this. I agree that using the XML parser
and xpath is probably nice
On 11/15/12 00:36, Eric Blake wrote:
As we enable more modes of snapshot creation, it becomes more important
to be able to quickly filter based on snapshot properties. This patch
introduces new filter flags; subsequent patches will introduce virsh
back-compat filtering, as well as actual libvirt
On 2012年11月19日 17:30, Osier Yang wrote:
Hi,
This proposal is trying to figure out a solution for migration
of domain which uses LUN behind vHBA as disk device (QEMU
emulated disk only at this stage). And other related NPIV
improvements which are not related with migration. I'm not
luck to get a
Hi,
This proposal is trying to figure out a solution for migration
of domain which uses LUN behind vHBA as disk device (QEMU
emulated disk only at this stage). And other related NPIV
improvements which are not related with migration. I'm not
luck to get a environment to test if the thoughts are w
Il 19/11/2012 04:45, Osier Yang ha scritto:
> On 2012年11月17日 06:47, Paolo Bonzini wrote:
>> Il 16/11/2012 10:59, Osier Yang ha scritto:
>>> If "path" for "ivshmem" memory device is specified, it's going
>>> to be setup with interrupts enabled. Otherwise interrupts is
>>> disabled. To be more clear,
49 matches
Mail list logo