[Openstack] Is HyperV support Flatdhcp using nova-network?

[Ray Sun]

Hi,
I try to manage HyperV network in nova-network using Flatdhcp, but seems it
not works, I wonder if nova-network can support HyperV in Flatdhcp model. I
always get a RPC timeout when getting network. Could anyone to help me with
the issues? Thanks.

Here's my configuration files:
Controller nova.conf:
# Network settings
network_manager=nova.network.manager.FlatDHCPManager
#network_manager=nova.network.manager.FlatManager
force_dhcp_release=True
dhcpbridge_flagfile=/etc/nova/nova.conf
dhcpbridge=/usr/bin/nova-dhcpbridge
firewall_driver=nova.virt.libvirt.firewall.IptablesFirewallDriver
public_interface=eth1
flat_interface=eth0
flat_network_bridge=br100
fixed_range=10.0.0.0/24
network_size=256
flat_injected=False
connection_type=libvirt
multi_host=True
libvirt_use_virtio_for_bridges=True
# Automatically adding floating IPs
auto_assign_floating_ip=False

HyperV nova.conf:
[DEFAULT]
auth_strategy=keystone
image_service=nova.image.glance.GlanceImageService
compute_driver=nova.virt.hyperv.driver.HyperVDriver
volume_api_class=nova.volume.cinder.API
fake_network=true
instances_path=C:\OpenStack\Instances
glance_api_servers=200.21.0.202:9292
use_cow_images=true
force_config_drive=false
injected_network_template=C:\Program Files (x86)\Cloudbase
Solutions\OpenStack\Nova\etc\interfaces.template
policy_file=C:\Program Files (x86)\Cloudbase
Solutions\OpenStack\Nova\etc\policy.json
mkisofs_cmd=C:\Program Files (x86)\Cloudbase
Solutions\OpenStack\Nova\bin\mkisofs.exe
verbose=true
allow_resize_to_same_host=true
running_deleted_instance_action=reap
running_deleted_instance_poll_interval=120
resize_confirm_window=5
resume_guests_state_on_host_boot=true
rpc_backend=nova.openstack.common.rpc.impl_kombu
rabbit_host=200.21.0.202
rabbit_port=5672
rabbit_userid=guest
logdir=C:\OpenStack\Log\
logfile=nova-compute.log
sql_connection=mysql://nova:novaPass@200.21.0.202/nova
[hyperv]
#vswitch_name=openstack
limit_cpu_features=false
config_drive_inject_password=false
qemu_img_cmd=C:\Program Files (x86)\Cloudbase
Solutions\OpenStack\Nova\bin\qemu-img.exe
config_drive_cdrom=true

Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] How to use Quantum with different hypervisers?

[Ray Sun]

Hello, All
I wonder if there's anyway to use Quantum with multiple hypervisors. For
example, in my cloud, I have KVM, VMWARE and HyperV, how can I make quantum
work for them, because they are using different quantum core plugins.

Thanks.

Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] OpenStack in China

[Ray Sun]

We met a lot of troubles when deploying OpenStack to CentOS 6.4, but we
want to stick to this.

Anyway, as I know, many companies use Ubuntu as their major OS for
Openstack.

Best Regards
-- Ray


On Fri, Aug 30, 2013 at 9:37 AM, pangj  wrote:

> On 2013-8-29 23:00, Kannan, Hari wrote:
>
>> Anyone know what are the top Linux distros for OpenStack in China?
>>
>
> I think Ubuntu is popular here.
> We have thousands of ubuntu servers, and we prepare to run a swift cluster
> right now.
>
> __**_
> Mailing list: http://lists.openstack.org/**cgi-bin/mailman/listinfo/**
> openstack 
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/**cgi-bin/mailman/listinfo/**
> openstack 
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] Keystone with memcache

[Ray Sun]

Stackers,
I have found there are two parts can set memcache in keystone.conf of Juno
release. I am not quite sure if they are duplicated or any difference?

Thanks.

[token]
driver = keystone.token.backends.memcache.Token
caching = True

[cache]
enabled = True
backend = dogpile.cache.memcached
backend_argument = url:localhost:11211



Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] OpenStack Kilo Release Note Translation

[Ray Sun]

Stackers,
I try to translate OpenStack Kilo release note, but I don't think I have
permission to do that even if I added translation tag for it. Any one who
have that permission, please help. Thanks.

Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] OpenStack Kilo Release Note Translation

[Ray Sun]

By the way, I found this in mediawiki translation plugin document:

This page contains changes which are not marked for translation." if you
are not a translation administrator

Best Regards
-- Ray

On Fri, May 1, 2015 at 10:38 AM, Ray Sun  wrote:

> Stackers,
> I try to translate OpenStack Kilo release note, but I don't think I have
> permission to do that even if I added translation tag for it. Any one who
> have that permission, please help. Thanks.
>
> Best Regards
> -- Ray
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] OpenStack Liberty Release Note Translation

[Ray Sun]

Stackers,
I try to translate OpenStack Liberty release note, but I don't think I have
permission to do that even if I added translation tag for it. Any one who
have that permission, please help. Thanks.

Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] OpenStack Kilo Release Note Translation

[Ray Sun]

I found this loop in previous release. Could somebody set the permission
for translation for Liberty release note?

Thanks.

Best Regards
-- Ray

On Fri, May 1, 2015 at 4:16 PM, Akira Yoshiyama 
wrote:

> Hi Tom,
>
> Thank you for preparing translation. In Japan, Kato-san just called
> volunteers for Japanese. I'll join them.
>
> On 2015年5月1日 12:01:50 GMT+09:00, Tom Fifield  wrote:
> >Hi,
> >
> >I just clicked "Mark This Page for Translation", which should fix this,
> >I think.
> >
> >"""
> >The page ReleaseNotes/Kilo has been marked up for translation with 166
> >translation units. The page can now be translated.
> >"""
> >
> >Regards,
> >
> >
> >Tom
> >
> >
> >On 01/05/15 10:40, Ray Sun wrote:
> >> By the way, I found this in mediawiki translation plugin document:
> >>
> >> This page contains changes which are not marked for translation." if
> >you
> >> are not a translation administrator
> >>
> >> Best Regards
> >> -- Ray
> >>
> >> On Fri, May 1, 2015 at 10:38 AM, Ray Sun  >> <mailto:xiaoq...@gmail.com>> wrote:
> >>
> >> Stackers,
> >> I try to translate OpenStack Kilo release note, but I don't think
> >I
> >> have permission to do that even if I added translation tag for
> >it.
> >> Any one who have that permission, please help. Thanks.
> >>
> >> Best Regards
> >> -- Ray
> >>
> >>
> >>
> >>
> >> ___
> >> Mailing list:
> >http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> >> Post to : openstack@lists.openstack.org
> >> Unsubscribe :
> >http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> >>
> >
> >
> >___
> >Mailing list:
> >http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> >Post to : openstack@lists.openstack.org
> >Unsubscribe :
> >http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>
> --
> Sent from my Android device with K-9 Mail. Please excuse my brevity.
>
> ___
> Mailing list:
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe :
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] [Neutron][HyperV]HyperV Networking Problem

[Ray Sun]

Dear all,
I am trying to test HyperV with OpenStack in Quantum VLAN model. But seems
the network can't be reachable.

*I use devstack to install my controller, by default, I use stable/grizzly
branch. Here's my localrc file:*
*disable_service n-net*
*enable_service q-svc*
*enable_service quantum*
*enable_service q-dhcp*
*enable_service q-l3*
*
*
*MULTI_HOST=True*
*
*
*PUBLIC_INTERFACE=eth1*
*FLOATING_RANGE=200.21.0.80/28*
*
*
*ENABLE_TENANT_VLANS=True*
*TENANT_VLAN_RANGE=1000:1999*
*PHYSICAL_NETWORK=default*
*OVS_PHYSICAL_BRIDGE=br-eth1*

*Then I create a new
file /etc/quantum/plugins/hyperv/hyperv_quantum_plugin.ini*
*[DATABASE]*
*sql_connection = mysql://root:admin@localhost/hyperv_quantum?charset=utf8*
*
*
*[HYPERV]*
*tenant_network_type = vlan*
*network_vlan_ranges = default:1000:2999*
*
*
*[SECURITYGROUP]*
*firewall_driver =
quantum.agent.linux.iptables_firewall.OVSHybridIptablesFirewallDriver*

*To avoid security groups problem, I also update nova.conf:*
*firewall_driver = nova.virt.libvirt.firewall.IptablesFirewallDriver
*
*
*
*The I restart nova-api service and modify quantum server command line,
using:*
*cd /opt/stack/neutron && python /opt/stack/neutron/bin/quantum-server
--config-file /etc/quantum/quantum.conf --confifile
/etc/quantum/plugins/hyperv/hyperv_quantum_plugin.ini
*
*
*
*Everything seems ok except these logs:*
2013-10-10 18:09:02ERROR [quantum.openstack.common.rpc.amqp] Exception
during message handling
Traceback (most recent call last):
  File "/opt/stack/neutron/quantum/openstack/common/rpc/amqp.py", line 430,
in _process_data
rval = self.proxy.dispatch(ctxt, version, method, **args)
  File "/opt/stack/neutron/quantum/common/rpc.py", line 43, in dispatch
quantum_ctxt, version, method, **kwargs)
  File "/opt/stack/neutron/quantum/openstack/common/rpc/dispatcher.py",
line 136, in dispatch
raise AttributeError("No such RPC function '%s'" % method)
AttributeError: No such RPC function 'report_state'

*But I don't think this will impact me, I remember I saw this bug on the
lanchpad, I follow the most introduction of the document:*
*http://www.cloudbase.it/quantum-hyper-v-plugin/*

After I finished all the steps, I tried to create a new instance using a
private network. I got my VM(Ubuntu) ready but without network connection.
As the pictures shown in the document above, I also got an external NIC
with VLAN id 1001 for the new VM.

Seems DHCP can not assign ip for the new VM.

*I try to tcpdump on my controller of my new VM's mac address, and I got:*
*0.0.0.0.bootpc > 255.255.255.255.bootps: [udp sum ok] BOOTP/DHCP,
Request from fa:16:3e:25:7a:73 (oui Unknown), length 300, xid 0x880a075d,
Flags [none] (0x)*
*  Client-Ethernet-Address fa:16:3e:25:7a:73 (oui Unknown)*
*  source link-address option (1), length 8 (1): fa:16:3e:25:7a:73*

But in my syslog of controller, I can't see any dhcp related logs.

My questions is:
1. In my local network we have 3 VLANs(id is 1, 2, 3), do I need to set new
VLANs based my configuration for quantum(1000:1999) on my hardware?
2. Do I need to setup any trunk port in my connection ports?
3. Can I use any other network type like Flatdhcp, even or nova-network for
simple?

Thanks.

Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [Neutron][HyperV]HyperV Networking Problem

[Ray Sun]

Pilotti,
Thanks for you quick response. I double check the value in my register,
seems all the related set to 0. But I still run the power shell script.
Then I restart my window server but seems it's still not works for me.

Then I did another test, I create a new VM on Hyper-V directly, then I set
the NIC to external and set the VLAN ID to 3(The same VLAN of the VM's
host), the VM still can not get any ip address.

How about any other reason for this problem? Thanks.

Best Regards
-- Ray


On Thu, Oct 10, 2013 at 9:53 PM, Alessandro Pilotti <
apilo...@cloudbasesolutions.com> wrote:

> Hi Ray,
>
>  Some Windows network adapter drivers disable VLAN access by default.
>
>  Here's Powershell script that will apply the adapter's proper
> configuration:
>
> https://bitbucket.org/cloudbase/vlanfilter-check/raw/8eea85e5cb38ce6e8ccb80fc00dd6c3f13869b6d/modifyRegistry.ps1
>
>  In case the script reports that one of morte adapters have been
> reconfigured, you'll need to reboot or restart the corresponding adapters.
>
>  If you prefer to perform the steps manually, open the following registry
> key:
>
>
> HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Class\{4d36e972-e325-11ce-bfc1-08002be10318}
>
> Look in all the child keys , e.g. 0001, 0002 etc for a value
> “VLanFiltering”. Make sure that if present it’s set to 0.
> In case of changes, reboot the server or restart the corresponding
> adapters.
>
>  Please let me know if this works for you.
>
>
>
>  Note: the error that you are reporting:
>
>   ERROR [quantum.openstack.common.rpc.amqp] Exception during message
> handling
>
>  has been fixed with https://review.openstack.org/#/c/46989/
>
>
>
>  On Oct 10, 2013, at 13:36 , Ray Sun  wrote:
>
>  Dear all,
> I am trying to test HyperV with OpenStack in Quantum VLAN model. But seems
> the network can't be reachable.
>
>  *I use devstack to install my controller, by default, I use
> stable/grizzly branch. Here's my localrc file:*
>  *disable_service n-net*
> *enable_service q-svc*
> *enable_service quantum*
> *enable_service q-dhcp*
> *enable_service q-l3*
> *
> *
>  *MULTI_HOST=True*
> *
> *
> *PUBLIC_INTERFACE=eth1*
> *FLOATING_RANGE=200.21.0.80/28*
> *
> *
>  *ENABLE_TENANT_VLANS=True*
> *TENANT_VLAN_RANGE=1000:1999*
> *PHYSICAL_NETWORK=default*
> *OVS_PHYSICAL_BRIDGE=br-eth1*
>
>  *Then I create a new
> file /etc/quantum/plugins/hyperv/hyperv_quantum_plugin.ini*
>  *[DATABASE]*
> *sql_connection = mysql://root:admin@localhost/hyperv_quantum?charset=utf8
> *
> *
> *
> *[HYPERV]*
> *tenant_network_type = vlan*
> *network_vlan_ranges = default:1000:2999*
> *
> *
> *[SECURITYGROUP]*
> *firewall_driver =
> quantum.agent.linux.iptables_firewall.OVSHybridIptablesFirewallDriver*
>
>  *To avoid security groups problem, I also update nova.conf:*
> *firewall_driver = nova.virt.libvirt.firewall.IptablesFirewallDriver
> *
> *
> *
> *The I restart nova-api service and modify quantum server command line,
> using:*
> *cd /opt/stack/neutron && python /opt/stack/neutron/bin/quantum-server
> --config-file /etc/quantum/quantum.conf --confifile
> /etc/quantum/plugins/hyperv/hyperv_quantum_plugin.ini
> *
> *
> *
> *Everything seems ok except these logs:*
>  2013-10-10 18:09:02ERROR [quantum.openstack.common.rpc.amqp]
> Exception during message handling
> Traceback (most recent call last):
>   File "/opt/stack/neutron/quantum/openstack/common/rpc/amqp.py", line
> 430, in _process_data
> rval = self.proxy.dispatch(ctxt, version, method, **args)
>   File "/opt/stack/neutron/quantum/common/rpc.py", line 43, in dispatch
> quantum_ctxt, version, method, **kwargs)
>   File "/opt/stack/neutron/quantum/openstack/common/rpc/dispatcher.py",
> line 136, in dispatch
> raise AttributeError("No such RPC function '%s'" % method)
> AttributeError: No such RPC function 'report_state'
>
>  *But I don't think this will impact me, I remember I saw this bug on the
> lanchpad, I follow the most introduction of the document:*
> *http://www.cloudbase.it/quantum-hyper-v-plugin/*
>
>  After I finished all the steps, I tried to create a new instance using a
> private network. I got my VM(Ubuntu) ready but without network connection.
> As the pictures shown in the document above, I also got an external NIC
> with VLAN id 1001 for the new VM.
>
>  Seems DHCP can not assign ip for the new VM.
>
>  *I try to tcpdump on my controller of my new VM's mac address, and I got:
> *
>  *0.0.0.0.bootpc > 255.255.255.255.bootps: [udp sum ok] BOOTP/DHCP,
> Request from 

[Openstack] [Nova]How to setup shared storage for live migration based on IPSAN device?

[Ray Sun]

Hi Stackers,
I found there is a document to describe how to setup live migration using
NFS by Mirantis.
http://www.mirantis.com/blog/tutorial-openstack-live-migration-with-kvm-hypervisor-and-nfs-shared-storage
In this articles, there is a paragraph here:
*In a typical Openstack deployment, every compute node manages its
instances locally in a dedicated directory (for example,
/var/lib/nova/instances/) but for live migration, this folder has to be in
a centralized location and shared across all the compute nodes. Hence, a
shared file system or block storage is an important requirement for
enabling live migration. For shared storage, a distributed file system such
as GlusterFS, NFS needs to be properly configured and running before live
migration can be performed. SAN storage protocols such as Fibre Channel
(FC) and iSCSI can also be used for shared storage.*

Please pay attention the sentence in blue, currently, I have a IPSAN
device, and I want use it as the shared storage for all my compute nodes.
My questions is how to setup it in nova, I didn't find any document to
introduce this, please help. Thanks a lot.

Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [Nova]How to setup shared storage for live migration based on IPSAN device?

[Ray Sun]

Daniel,
Thanks for you response. But I am still confusing about use a iscsi device,
if I want to use it as shared storage, first I need to attach to a node as
a block storage, then I mount it to other compute nodes using NFS. The
problem is that this will make a big lost on performance. Currently nova
already support to create a new VM on ceph RBD, but why we can not create
it on a block storage?

Thanks.

在 2013年10月28日星期一，Speichert,Daniel 写道：

>  Ray,
>
> It seems that OpenStack purposefully omits the method of provisioning
> shared storage.
>
> It’s totally up to you �C all you need to do is to mount
> /var/lib/nova/instances/ on each compute node on the system level.
>
> ** **
>
> Make sure your user IDs inside /var/lib/nova/instances/ are correct: nova,
> libvirt-em and kvm need to have matching IDs across all machines if you use
> e.g. NFSv3.
>
> ** **
>
> Regards,
>
> Daniel
>
> ** **
>
> *From:* Ray Sun [mailto:xiaoq...@gmail.com  'xiaoq...@gmail.com');>]
> *Sent:* Monday, October 28, 2013 4:27 AM
> *To:* OpenStack Maillist
> *Subject:* [Openstack] [Nova]How to setup shared storage for live
> migration based on IPSAN device?
>
> ** **
>
> Hi Stackers,
>
> I found there is a document to describe how to setup live migration using
> NFS by Mirantis.
>
>
> http://www.mirantis.com/blog/tutorial-openstack-live-migration-with-kvm-hypervisor-and-nfs-shared-storage
> 
>
> In this articles, there is a paragraph here:
>
> *In a typical Openstack deployment, every compute node manages its
> instances locally in a dedicated directory (for example,
> /var/lib/nova/instances/) but for live migration, this folder has to be in
> a centralized location and shared across all the compute nodes. Hence, a
> shared file system or block storage is an important requirement for
> enabling live migration. For shared storage, a distributed file system such
> as GlusterFS, NFS needs to be properly configured and running before live
> migration can be performed. SAN storage protocols such as Fibre Channel
> (FC) and iSCSI can also be used for shared storage.*
>
> ** **
>
> Please pay attention the sentence in blue, currently, I have a IPSAN
> device, and I want use it as the shared storage for all my compute nodes.
> My questions is how to setup it in nova, I didn't find any document to
> introduce this, please help. Thanks a lot.
>
> ** **
>
> Best Regards
> -- Ray
>


-- 
Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [Nova]How to setup shared storage for live migration based on IPSAN device?

[Ray Sun]

Thanks for all your response.

So libvirt can support ISCSI LUN, but OpenStack didn't provide such options
yet?

Best Regards
-- Ray


On Tue, Oct 29, 2013 at 1:11 AM, Razique Mahroua
wrote:

> The only thing is that you need to create the volumes on your ISCSI
> backend first, so libvirt can use them, otherwise as a shared storage,
> works fine. ISCSI always gave me nice speeds compared to NFS
>
> On Oct 28, 2013, at 10:04, Chris Friesen 
> wrote:
>
> > That works for one client.  How do you synchronize access between
> multiple clients?
> >
> > Also, for iSCSI it looks like libvirt can't create/delete volumes, that
> needs to be done on the server.
> >
> > Chris
> >
> > On 10/28/2013 10:49 AM, Razique Mahroua wrote:
> >> Libvrit does support ISCSI LUN as backends, meaning you can mount the
> ISCSI block on every compute node and you should be fine
> >> - Razique
> >>
> >> On Oct 28, 2013, at 8:54, Chris Friesen 
> wrote:
> >>
> >>> On 10/28/2013 09:35 AM, Ray Sun wrote:
> >>>> Daniel,
> >>>> Thanks for you response. But I am still confusing about use a iscsi
> >>>> device, if I want to use it as shared storage, first I need to attach
> to
> >>>> a node as a block storage, then I mount it to other compute nodes
> using
> >>>> NFS. The problem is that this will make a big lost on performance.
> >>>> Currently nova already support to create a new VM on ceph RBD, but why
> >>>> we can not create it on a block storage?
> >>>
> >>> ceph and NFS are both designed for simultaneous access by multiple
> clients.
> >>>
> >>> iSCSI is not.
> >>>
> >>> Chris
> >>>
> >>>
> >>> ___
> >>> Mailing list:
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> >>> Post to : openstack@lists.openstack.org
> >>> Unsubscribe :
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> >>
> >
>
>
> ___
> Mailing list:
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe :
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [Nova]How to setup shared storage for live migration based on IPSAN device?

[Ray Sun]

So is there any blueprint for this feature?

Best Regards
-- Ray


On Tue, Oct 29, 2013 at 1:48 PM, Razique Mahroua
wrote:

> That’s is, OpenStack is not aware yet of the backend that servers the
> /var/lib/nova/instances directory ; not that I’m aware of :)
> - Razique
>
> On Oct 28, 2013, at 22:35, Ray Sun  wrote:
>
> Thanks for all your response.
>
> So libvirt can support ISCSI LUN, but OpenStack didn't provide such
> options yet?
>
> Best Regards
> -- Ray
>
>
> On Tue, Oct 29, 2013 at 1:11 AM, Razique Mahroua <
> razique.mahr...@gmail.com> wrote:
>
>> The only thing is that you need to create the volumes on your ISCSI
>> backend first, so libvirt can use them, otherwise as a shared storage,
>> works fine. ISCSI always gave me nice speeds compared to NFS
>>
>> On Oct 28, 2013, at 10:04, Chris Friesen 
>> wrote:
>>
>> > That works for one client.  How do you synchronize access between
>> multiple clients?
>> >
>> > Also, for iSCSI it looks like libvirt can't create/delete volumes, that
>> needs to be done on the server.
>> >
>> > Chris
>> >
>> > On 10/28/2013 10:49 AM, Razique Mahroua wrote:
>> >> Libvrit does support ISCSI LUN as backends, meaning you can mount the
>> ISCSI block on every compute node and you should be fine
>> >> - Razique
>> >>
>> >> On Oct 28, 2013, at 8:54, Chris Friesen 
>> wrote:
>> >>
>> >>> On 10/28/2013 09:35 AM, Ray Sun wrote:
>> >>>> Daniel,
>> >>>> Thanks for you response. But I am still confusing about use a iscsi
>> >>>> device, if I want to use it as shared storage, first I need to
>> attach to
>> >>>> a node as a block storage, then I mount it to other compute nodes
>> using
>> >>>> NFS. The problem is that this will make a big lost on performance.
>> >>>> Currently nova already support to create a new VM on ceph RBD, but
>> why
>> >>>> we can not create it on a block storage?
>> >>>
>> >>> ceph and NFS are both designed for simultaneous access by multiple
>> clients.
>> >>>
>> >>> iSCSI is not.
>> >>>
>> >>> Chris
>> >>>
>> >>>
>> >>> ___
>> >>> Mailing list:
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>> >>> Post to : openstack@lists.openstack.org
>> >>> Unsubscribe :
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>> >>
>> >
>>
>>
>> ___
>> Mailing list:
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>> Post to : openstack@lists.openstack.org
>> Unsubscribe :
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>
>>
>
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] How to support for multiple external networks?

[Ray Sun]

I want to setup multiple external networks, but I don't know how to make
this works.

Here's my user requirements:
There are two departments want to use my cloud, but as history problem,
they wants their VMs to use original IP range. DepA wants to use floating
ip in 192.168.66.0/24, DepB wants to use floating ip in 192.168.77.0/24.

I can create multiple subnets in my external network. But I am not quite
sure if this can work in OpenStack. In my network environment,
192.168.66.0/24 and 192.168.77.0/24 are two VLANs.


Anybody know how to implement this kind of scenarios? If this is not
possible, how can I design my architecture to meet the requirements.

Thanks.

Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] How to support for multiple external networks?

[Ray Sun]

Hauke,
This is interesting, Is there any way to specify tenantA can only use
ext-subnet1 and tenantB only use ext-subnet2?

Thanks.

Best Regards
-- Ray

On Thu, Nov 19, 2015 at 7:27 PM, Hauke Bruno Wollentin <
hauke-bruno.wollen...@innovo-cloud.de> wrote:

> Hi Ray,
>
> in general it is possible to use two (or more) subnets in one external
> network, but all instances then will get IPs from both subnets (if the
> floating IPs on subnetA are allocated, the next ones will be allocated from
> subnetB).
>
> So you need to create two (non-shared, tenant specific) external networks.
> For
> the VLAN dedication you could use 1 L3 Agent per 1 external network.
>
> cheers,
> hauke
>
> ---
> original message
> timestamp: Thursday, November 19, 2015 06:18:06 PM
> from: Ray Sun 
> to: openstack@lists.openstack.org
> cc:
> subject: [Openstack] How to support for multiple external networks?
> message id:  c...@mail.gmail.com>
>
> > I want to setup multiple external networks, but I don't know how to make
> > this works.
> >
> > Here's my user requirements:
> > There are two departments want to use my cloud, but as history problem,
> > they wants their VMs to use original IP range. DepA wants to use floating
> > ip in 192.168.66.0/24, DepB wants to use floating ip in 192.168.77.0/24.
> >
> > I can create multiple subnets in my external network. But I am not quite
> > sure if this can work in OpenStack. In my network environment,
> > 192.168.66.0/24 and 192.168.77.0/24 are two VLANs.
> >
> >
> > Anybody know how to implement this kind of scenarios? If this is not
> > possible, how can I design my architecture to meet the requirements.
> >
> > Thanks.
> >
> > Best Regards
> > -- Ray
> --
> Hauke Bruno Wollentin
> (Infrastructure Engineer Cloud)
> ___
> Mailing list:
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe :
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] How to support for multiple external networks?

[Ray Sun]

​Thanks, I will have a try.​

Best Regards
-- Ray

On Thu, Nov 19, 2015 at 7:38 PM, José Riguera López 
wrote:

>
> You can, just tell neutron you want to have another external network. Lets
> say VLAN=300 is 192.168.66. <http://192.168.66.0/24>0/24
>
> # Creating the external network type VLAN with VLAND ID == 300
> # The name of the provider physical_network "ext" is mapped to a physical
> # device on the ML2 configuration file of each network node
>
> neutron net-create \
> --provider:network_type vlan \
> --provider:physical_network ext \
> --router:external \
> --provider:segmentation_id 300 \
> ext_net_300
>
> # Define the IP range for floating IPs available on that network, the
> gateway,
> # DNS nameservers and CIDR, and no DHCP (in this case)
> neutron subnet-create \
> --allocation-pool start=192.168.66. <http://192.168.66.0/24>10,end=1
> 92.168.66. <http://192.168.66.0/24>255 \
> --ip-version 4 \
> --gateway 192.168.88.1 \
> --disable-dhcp \
> --dns-nameserver 8.8.8.8 \
> --name ext_net_300_subnet \
> ext_net_300 192.168.66. <http://192.168.66.0/24>0/24
>
> Then the users will see 2 (or more) floating ip pools ...
>
> Regards,
>
>
>
> 2015-11-19 11:18 GMT+01:00 Ray Sun :
>
>> I want to setup multiple external networks, but I don't know how to make
>> this works.
>>
>> Here's my user requirements:
>> There are two departments want to use my cloud, but as history problem,
>> they wants their VMs to use original IP range. DepA wants to use floating
>> ip in 192.168.66.0/24, DepB wants to use floating ip in 192.168.77.0/24.
>>
>> I can create multiple subnets in my external network. But I am not quite
>> sure if this can work in OpenStack. In my network environment,
>> 192.168.66.0/24 and 192.168.77.0/24 are two VLANs.
>>
>>
>> Anybody know how to implement this kind of scenarios? If this is not
>> possible, how can I design my architecture to meet the requirements.
>>
>> Thanks.
>>
>> Best Regards
>> -- Ray
>>
>> ___
>> Mailing list:
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>> Post to : openstack@lists.openstack.org
>> Unsubscribe :
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>
>>
>
>
> --
> José Riguera López 
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] Upload image to specific cinder backend

[Ray Sun]

Currently, we use cinder as glance backend and all our instances can boot
from volume. But when I have multiple cinder backends, example: EMC and
ceph, I can only specific one cinder_volume_backend in the glance
configuration.

My question: Is there anyway to specific the image I want to upload in API
level? Or any other way I can control it.

Thanks.

Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] [OpenStack][Kolla]Choose binary or source

[Ray Sun]

OpenStackers,
I have several questions for using kolla deployment in production:
1. Binary or source? Which one is mostly used in production
2. Which one is easy to operation, like updating.
3. Is there any document to describe how to operation kolla deployment
OpenStack, seems it's not the quite same with the bare metal deployment.

Thanks.

Best Regards
-- Ray
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [OpenStack][Kolla]Choose binary or source

[Ray Sun]

Eric,
Thanks for your reply.

I just consider one senario: if I deployed use source, if source is
updated, do I need to rebuild the docker image and do an update? Or is
there any better way?

Best Regards
-- Ray


On Mon, Jun 18, 2018 at 10:22 AM Erik McCormick 
wrote:

>
>
> On Sun, Jun 17, 2018, 8:46 PM Ray Sun  wrote:
>
>> OpenStackers,
>> I have several questions for using kolla deployment in production:
>>
> 1. Binary or source? Which one is mostly used in production
>>
>
> I've heard a lot more reports from people running source, but this is
> anecdotal
>
> 2. Which one is easy to operation, like updating..
>>
>
> IMHO source is better for updates. With binary you sit around waiting for
> packages to decide a patch or stable milestone is worthy of dropping new
> packages. With source you can get things when they hit stable branches
> upstream. I did exactly one binary deployment, waited forever for a patch I
> needed, and swore off doing it that way forever more.
>
> Operations are the same either way really.
>
> ...
>> 3. Is there any document to describe how to operation kolla deployment
>> OpenStack, seems it's not the quite same with the bare metal deployment.
>>
>
> The deployment docs are quite good. There isn't really a good ops guide
> though. This is true of bare metal as well though ( in process of being
> addressed presently). The IRC channel is terrific though, so if you hit a
> wall, pop on there and ask for help or email this list with a tag like you
> did here.
>
>>
>> Thanks.
>>
>> Best Regards
>> -- Ray
>>
>
> Cherrs,
> Erik
>
> ___
>> Mailing list:
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>> Post to : openstack@lists.openstack.org
>> Unsubscribe :
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [OpenStack][Kolla]Choose binary or source

[Ray Sun]

Ok, thanks for the info.
Best Regards
-- Ray


On Mon, Jun 18, 2018 at 11:35 AM Erik McCormick 
wrote:

>
>
> On Sun, Jun 17, 2018, 10:47 PM Ray Sun  wrote:
>
>> Eric,
>> Thanks for your reply.
>>
>> I just consider one senario: if I deployed use source, if source is
>> updated, do I need to rebuild the docker image and do an update? Or is
>> there any better way?
>>
>> Best Regards
>> -- Ray
>>
>
> If you want the updates, then build new containers with new tags and run
> the upgrade playbook. You can do this even just for one project.
>
> The process is the same for binary or source. If you want updates, build
> new containers with newer tags. You don't upgrade the already-running
> containers in place.
>
> -Erik
>
>
>>
>> On Mon, Jun 18, 2018 at 10:22 AM Erik McCormick <
>> emccorm...@cirrusseven.com> wrote:
>>
>>>
>>>
>>> On Sun, Jun 17, 2018, 8:46 PM Ray Sun  wrote:
>>>
>>>> OpenStackers,
>>>> I have several questions for using kolla deployment in production:
>>>>
>>> 1. Binary or source? Which one is mostly used in production
>>>>
>>>
>>> I've heard a lot more reports from people running source, but this is
>>> anecdotal
>>>
>>> 2. Which one is easy to operation, like updating..
>>>>
>>>
>>> IMHO source is better for updates. With binary you sit around waiting
>>> for packages to decide a patch or stable milestone is worthy of dropping
>>> new packages. With source you can get things when they hit stable branches
>>> upstream. I did exactly one binary deployment, waited forever for a patch I
>>> needed, and swore off doing it that way forever more.
>>>
>>> Operations are the same either way really.
>>>
>>> ...
>>>> 3. Is there any document to describe how to operation kolla deployment
>>>> OpenStack, seems it's not the quite same with the bare metal deployment.
>>>>
>>>
>>> The deployment docs are quite good. There isn't really a good ops guide
>>> though. This is true of bare metal as well though ( in process of being
>>> addressed presently). The IRC channel is terrific though, so if you hit a
>>> wall, pop on there and ask for help or email this list with a tag like you
>>> did here.
>>>
>>>>
>>>> Thanks.
>>>>
>>>> Best Regards
>>>> -- Ray
>>>>
>>>
>>> Cherrs,
>>> Erik
>>>
>>> ___
>>>> Mailing list:
>>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>>> Post to : openstack@lists.openstack.org
>>>> Unsubscribe :
>>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>>>
>>>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [OpenStack][Kolla]Choose binary or source

[Ray Sun]

Eric,
I discussed upgrade action today with my colleagues, our purpose is a
little difference. We want to create an ISO with kolla images and help our
customer install OpenStack in an offline way. But for upgrade each time,
user's environment always can't access to the internet, and each time I
need to copy the docker exported file, it's very large and it seems it's
not worth to update one file but copy a 1G file.

I am not quite sure there is other way to handle this scenarios.


Thanks.
Best Regards
-- Ray


On Mon, Jun 18, 2018 at 12:14 PM Ray Sun  wrote:

> Ok, thanks for the info.
> Best Regards
> -- Ray
>
>
> On Mon, Jun 18, 2018 at 11:35 AM Erik McCormick <
> emccorm...@cirrusseven.com> wrote:
>
>>
>>
>> On Sun, Jun 17, 2018, 10:47 PM Ray Sun  wrote:
>>
>>> Eric,
>>> Thanks for your reply.
>>>
>>> I just consider one senario: if I deployed use source, if source is
>>> updated, do I need to rebuild the docker image and do an update? Or is
>>> there any better way?
>>>
>>> Best Regards
>>> -- Ray
>>>
>>
>> If you want the updates, then build new containers with new tags and run
>> the upgrade playbook. You can do this even just for one project.
>>
>> The process is the same for binary or source. If you want updates, build
>> new containers with newer tags. You don't upgrade the already-running
>> containers in place.
>>
>> -Erik
>>
>>
>>>
>>> On Mon, Jun 18, 2018 at 10:22 AM Erik McCormick <
>>> emccorm...@cirrusseven.com> wrote:
>>>
>>>>
>>>>
>>>> On Sun, Jun 17, 2018, 8:46 PM Ray Sun  wrote:
>>>>
>>>>> OpenStackers,
>>>>> I have several questions for using kolla deployment in production:
>>>>>
>>>> 1. Binary or source? Which one is mostly used in production
>>>>>
>>>>
>>>> I've heard a lot more reports from people running source, but this is
>>>> anecdotal
>>>>
>>>> 2. Which one is easy to operation, like updating..
>>>>>
>>>>
>>>> IMHO source is better for updates. With binary you sit around waiting
>>>> for packages to decide a patch or stable milestone is worthy of dropping
>>>> new packages. With source you can get things when they hit stable branches
>>>> upstream. I did exactly one binary deployment, waited forever for a patch I
>>>> needed, and swore off doing it that way forever more.
>>>>
>>>> Operations are the same either way really.
>>>>
>>>> ...
>>>>> 3. Is there any document to describe how to operation kolla deployment
>>>>> OpenStack, seems it's not the quite same with the bare metal deployment.
>>>>>
>>>>
>>>> The deployment docs are quite good. There isn't really a good ops guide
>>>> though. This is true of bare metal as well though ( in process of being
>>>> addressed presently). The IRC channel is terrific though, so if you hit a
>>>> wall, pop on there and ask for help or email this list with a tag like you
>>>> did here.
>>>>
>>>>>
>>>>> Thanks.
>>>>>
>>>>> Best Regards
>>>>> -- Ray
>>>>>
>>>>
>>>> Cherrs,
>>>> Erik
>>>>
>>>> ___
>>>>> Mailing list:
>>>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>>>> Post to : openstack@lists.openstack.org
>>>>> Unsubscribe :
>>>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>>>>
>>>>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack