You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@cloudstack.apache.org by Bryan Whitehead <dr...@megahappy.net> on 2013/12/09 21:01:55 UTC
duplicate system UUID's with KVM
I have 3 independent Cloudstack installs. One is 3.0.x and the others are
4.1.0.
Using KVM (i'm only using KVM so I don't have anything else for
comparison), between 3.0.x and 4.1.0 I'm getting instances with UUID's that
are the same.
I get the UUID by running this on the console (CentOS):
dmidecode -s system-uuid
Here is example output from 1 host:
[root@fortress ~]# dmidecode -s system-uuid
C1260F04-F171-3136-85A7-F0B77699DA33
[root@fortress ~]# ifconfig eth0
eth0 Link encap:Ethernet HWaddr 06:9A:36:00:00:AF
inet addr:removed Bcast:70.33.251.255 Mask:255.255.255.0
inet6 addr: fe80::49a:36ff:fe00:af/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:20586 errors:0 dropped:0 overruns:0 frame:0
TX packets:1796 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:1764556 (1.6 MiB) TX bytes:197329 (192.7 KiB)
here is another:
[root@db-sla01 ~]# dmidecode -s system-uuid
C1260F04-F171-3136-85A7-F0B77699DA33
[root@db-sla01 ~]# ifconfig eth0
eth0 Link encap:Ethernet HWaddr 06:5E:9A:00:00:C9
inet addr:removed Bcast:64.13.168.255 Mask:255.255.255.0
inet6 addr: fe80::45e:9aff:fe00:c9/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:7644414 errors:0 dropped:0 overruns:0 frame:0
TX packets:3073765 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:735505477 (701.4 MiB) TX bytes:519743789 (495.6 MiB)
NOTE: The IP's are not in the same nor in the same subnet
The time between creation is pretty long... weeks.
Any ideas?
I've just been killing VM's that have collisions with UUID's but it happens
pretty often.
-Bryan
Re: duplicate system UUID's with KVM
Posted by Chiradeep Vittal <Ch...@citrix.com>.
It should be the same as the VM UUID, which is a pseudo-random UUID.
On 12/10/13 10:59 AM, "Bryan Whitehead" <dr...@megahappy.net> wrote:
>(Sorry, meant to say Cloudstack 3.0.x boxes are running CentOS6.2
>still...)
>
>
>On Tue, Dec 10, 2013 at 10:59 AM, Bryan Whitehead
><dr...@megahappy.net>wrote:
>
>> In all cases these are CentOS boxes. The 3.0.x boxes are still in
>> CentOS6.x land but the 4.1 Cloudstack boxes are 6.4+updates (as of
>>6months
>> ago).
>>
>> I don't know if the UUID internal to the VM is generated by cloudstack,
>> libvirtd, or qemu-kvm. Since the mac addresses have never had a
>>collision I
>> suspect the UUID is random with a common seed. Just not sure what piece
>>is
>> doing creating the UUID for a fresh VM.
>>
>>
>> On Mon, Dec 9, 2013 at 10:04 PM, Chiradeep Vittal <
>> Chiradeep.Vittal@citrix.com> wrote:
>>
>>> What is the OS of the KVM host?
>>> I believe vm uuids are type 4 uuids and are hence independent of time.
>>> http://docs.oracle.com/javase/6/docs/api/java/util/UUID.html
>>>
>>>
>>>
>>> On 12/9/13 12:01 PM, "Bryan Whitehead" <dr...@megahappy.net> wrote:
>>>
>>> >I have 3 independent Cloudstack installs. One is 3.0.x and the others
>>>are
>>> >4.1.0.
>>> >
>>> >Using KVM (i'm only using KVM so I don't have anything else for
>>> >comparison), between 3.0.x and 4.1.0 I'm getting instances with UUID's
>>> >that
>>> >are the same.
>>> >
>>> >I get the UUID by running this on the console (CentOS):
>>> >
>>> >dmidecode -s system-uuid
>>> >
>>> >Here is example output from 1 host:
>>> >[root@fortress ~]# dmidecode -s system-uuid
>>> >C1260F04-F171-3136-85A7-F0B77699DA33
>>> >[root@fortress ~]# ifconfig eth0
>>> >eth0 Link encap:Ethernet HWaddr 06:9A:36:00:00:AF
>>> > inet addr:removed Bcast:70.33.251.255 Mask:255.255.255.0
>>> > inet6 addr: fe80::49a:36ff:fe00:af/64 Scope:Link
>>> > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
>>> > RX packets:20586 errors:0 dropped:0 overruns:0 frame:0
>>> > TX packets:1796 errors:0 dropped:0 overruns:0 carrier:0
>>> > collisions:0 txqueuelen:1000
>>> > RX bytes:1764556 (1.6 MiB) TX bytes:197329 (192.7 KiB)
>>> >
>>> >
>>> >here is another:
>>> >[root@db-sla01 ~]# dmidecode -s system-uuid
>>> >C1260F04-F171-3136-85A7-F0B77699DA33
>>> >[root@db-sla01 ~]# ifconfig eth0
>>> >eth0 Link encap:Ethernet HWaddr 06:5E:9A:00:00:C9
>>> > inet addr:removed Bcast:64.13.168.255 Mask:255.255.255.0
>>> > inet6 addr: fe80::45e:9aff:fe00:c9/64 Scope:Link
>>> > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
>>> > RX packets:7644414 errors:0 dropped:0 overruns:0 frame:0
>>> > TX packets:3073765 errors:0 dropped:0 overruns:0 carrier:0
>>> > collisions:0 txqueuelen:1000
>>> > RX bytes:735505477 (701.4 MiB) TX bytes:519743789 (495.6
>>>MiB)
>>> >
>>> >NOTE: The IP's are not in the same nor in the same subnet
>>> >
>>> >The time between creation is pretty long... weeks.
>>> >
>>> >Any ideas?
>>> >
>>> >I've just been killing VM's that have collisions with UUID's but it
>>> >happens
>>> >pretty often.
>>> >
>>> >-Bryan
>>>
>>>
>>
Re: duplicate system UUID's with KVM
Posted by Bryan Whitehead <dr...@megahappy.net>.
(Sorry, meant to say Cloudstack 3.0.x boxes are running CentOS6.2 still...)
On Tue, Dec 10, 2013 at 10:59 AM, Bryan Whitehead <dr...@megahappy.net>wrote:
> In all cases these are CentOS boxes. The 3.0.x boxes are still in
> CentOS6.x land but the 4.1 Cloudstack boxes are 6.4+updates (as of 6months
> ago).
>
> I don't know if the UUID internal to the VM is generated by cloudstack,
> libvirtd, or qemu-kvm. Since the mac addresses have never had a collision I
> suspect the UUID is random with a common seed. Just not sure what piece is
> doing creating the UUID for a fresh VM.
>
>
> On Mon, Dec 9, 2013 at 10:04 PM, Chiradeep Vittal <
> Chiradeep.Vittal@citrix.com> wrote:
>
>> What is the OS of the KVM host?
>> I believe vm uuids are type 4 uuids and are hence independent of time.
>> http://docs.oracle.com/javase/6/docs/api/java/util/UUID.html
>>
>>
>>
>> On 12/9/13 12:01 PM, "Bryan Whitehead" <dr...@megahappy.net> wrote:
>>
>> >I have 3 independent Cloudstack installs. One is 3.0.x and the others are
>> >4.1.0.
>> >
>> >Using KVM (i'm only using KVM so I don't have anything else for
>> >comparison), between 3.0.x and 4.1.0 I'm getting instances with UUID's
>> >that
>> >are the same.
>> >
>> >I get the UUID by running this on the console (CentOS):
>> >
>> >dmidecode -s system-uuid
>> >
>> >Here is example output from 1 host:
>> >[root@fortress ~]# dmidecode -s system-uuid
>> >C1260F04-F171-3136-85A7-F0B77699DA33
>> >[root@fortress ~]# ifconfig eth0
>> >eth0 Link encap:Ethernet HWaddr 06:9A:36:00:00:AF
>> > inet addr:removed Bcast:70.33.251.255 Mask:255.255.255.0
>> > inet6 addr: fe80::49a:36ff:fe00:af/64 Scope:Link
>> > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
>> > RX packets:20586 errors:0 dropped:0 overruns:0 frame:0
>> > TX packets:1796 errors:0 dropped:0 overruns:0 carrier:0
>> > collisions:0 txqueuelen:1000
>> > RX bytes:1764556 (1.6 MiB) TX bytes:197329 (192.7 KiB)
>> >
>> >
>> >here is another:
>> >[root@db-sla01 ~]# dmidecode -s system-uuid
>> >C1260F04-F171-3136-85A7-F0B77699DA33
>> >[root@db-sla01 ~]# ifconfig eth0
>> >eth0 Link encap:Ethernet HWaddr 06:5E:9A:00:00:C9
>> > inet addr:removed Bcast:64.13.168.255 Mask:255.255.255.0
>> > inet6 addr: fe80::45e:9aff:fe00:c9/64 Scope:Link
>> > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
>> > RX packets:7644414 errors:0 dropped:0 overruns:0 frame:0
>> > TX packets:3073765 errors:0 dropped:0 overruns:0 carrier:0
>> > collisions:0 txqueuelen:1000
>> > RX bytes:735505477 (701.4 MiB) TX bytes:519743789 (495.6 MiB)
>> >
>> >NOTE: The IP's are not in the same nor in the same subnet
>> >
>> >The time between creation is pretty long... weeks.
>> >
>> >Any ideas?
>> >
>> >I've just been killing VM's that have collisions with UUID's but it
>> >happens
>> >pretty often.
>> >
>> >-Bryan
>>
>>
>
Re: duplicate system UUID's with KVM
Posted by Bryan Whitehead <dr...@megahappy.net>.
In all cases these are CentOS boxes. The 3.0.x boxes are still in CentOS6.x
land but the 4.1 Cloudstack boxes are 6.4+updates (as of 6months ago).
I don't know if the UUID internal to the VM is generated by cloudstack,
libvirtd, or qemu-kvm. Since the mac addresses have never had a collision I
suspect the UUID is random with a common seed. Just not sure what piece is
doing creating the UUID for a fresh VM.
On Mon, Dec 9, 2013 at 10:04 PM, Chiradeep Vittal <
Chiradeep.Vittal@citrix.com> wrote:
> What is the OS of the KVM host?
> I believe vm uuids are type 4 uuids and are hence independent of time.
> http://docs.oracle.com/javase/6/docs/api/java/util/UUID.html
>
>
>
> On 12/9/13 12:01 PM, "Bryan Whitehead" <dr...@megahappy.net> wrote:
>
> >I have 3 independent Cloudstack installs. One is 3.0.x and the others are
> >4.1.0.
> >
> >Using KVM (i'm only using KVM so I don't have anything else for
> >comparison), between 3.0.x and 4.1.0 I'm getting instances with UUID's
> >that
> >are the same.
> >
> >I get the UUID by running this on the console (CentOS):
> >
> >dmidecode -s system-uuid
> >
> >Here is example output from 1 host:
> >[root@fortress ~]# dmidecode -s system-uuid
> >C1260F04-F171-3136-85A7-F0B77699DA33
> >[root@fortress ~]# ifconfig eth0
> >eth0 Link encap:Ethernet HWaddr 06:9A:36:00:00:AF
> > inet addr:removed Bcast:70.33.251.255 Mask:255.255.255.0
> > inet6 addr: fe80::49a:36ff:fe00:af/64 Scope:Link
> > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> > RX packets:20586 errors:0 dropped:0 overruns:0 frame:0
> > TX packets:1796 errors:0 dropped:0 overruns:0 carrier:0
> > collisions:0 txqueuelen:1000
> > RX bytes:1764556 (1.6 MiB) TX bytes:197329 (192.7 KiB)
> >
> >
> >here is another:
> >[root@db-sla01 ~]# dmidecode -s system-uuid
> >C1260F04-F171-3136-85A7-F0B77699DA33
> >[root@db-sla01 ~]# ifconfig eth0
> >eth0 Link encap:Ethernet HWaddr 06:5E:9A:00:00:C9
> > inet addr:removed Bcast:64.13.168.255 Mask:255.255.255.0
> > inet6 addr: fe80::45e:9aff:fe00:c9/64 Scope:Link
> > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> > RX packets:7644414 errors:0 dropped:0 overruns:0 frame:0
> > TX packets:3073765 errors:0 dropped:0 overruns:0 carrier:0
> > collisions:0 txqueuelen:1000
> > RX bytes:735505477 (701.4 MiB) TX bytes:519743789 (495.6 MiB)
> >
> >NOTE: The IP's are not in the same nor in the same subnet
> >
> >The time between creation is pretty long... weeks.
> >
> >Any ideas?
> >
> >I've just been killing VM's that have collisions with UUID's but it
> >happens
> >pretty often.
> >
> >-Bryan
>
>
Re: duplicate system UUID's with KVM
Posted by Chiradeep Vittal <Ch...@citrix.com>.
What is the OS of the KVM host?
I believe vm uuids are type 4 uuids and are hence independent of time.
http://docs.oracle.com/javase/6/docs/api/java/util/UUID.html
On 12/9/13 12:01 PM, "Bryan Whitehead" <dr...@megahappy.net> wrote:
>I have 3 independent Cloudstack installs. One is 3.0.x and the others are
>4.1.0.
>
>Using KVM (i'm only using KVM so I don't have anything else for
>comparison), between 3.0.x and 4.1.0 I'm getting instances with UUID's
>that
>are the same.
>
>I get the UUID by running this on the console (CentOS):
>
>dmidecode -s system-uuid
>
>Here is example output from 1 host:
>[root@fortress ~]# dmidecode -s system-uuid
>C1260F04-F171-3136-85A7-F0B77699DA33
>[root@fortress ~]# ifconfig eth0
>eth0 Link encap:Ethernet HWaddr 06:9A:36:00:00:AF
> inet addr:removed Bcast:70.33.251.255 Mask:255.255.255.0
> inet6 addr: fe80::49a:36ff:fe00:af/64 Scope:Link
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:20586 errors:0 dropped:0 overruns:0 frame:0
> TX packets:1796 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:1000
> RX bytes:1764556 (1.6 MiB) TX bytes:197329 (192.7 KiB)
>
>
>here is another:
>[root@db-sla01 ~]# dmidecode -s system-uuid
>C1260F04-F171-3136-85A7-F0B77699DA33
>[root@db-sla01 ~]# ifconfig eth0
>eth0 Link encap:Ethernet HWaddr 06:5E:9A:00:00:C9
> inet addr:removed Bcast:64.13.168.255 Mask:255.255.255.0
> inet6 addr: fe80::45e:9aff:fe00:c9/64 Scope:Link
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:7644414 errors:0 dropped:0 overruns:0 frame:0
> TX packets:3073765 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:1000
> RX bytes:735505477 (701.4 MiB) TX bytes:519743789 (495.6 MiB)
>
>NOTE: The IP's are not in the same nor in the same subnet
>
>The time between creation is pretty long... weeks.
>
>Any ideas?
>
>I've just been killing VM's that have collisions with UUID's but it
>happens
>pretty often.
>
>-Bryan