[one-users] OCCI vm status indication
Daniel Molina
dmolina at opennebula.org
Mon Jun 10 02:44:06 PDT 2013
Hi,
On 7 June 2013 08:16, Miloš Kozák <milos.kozak at lejmr.com> wrote:
> Hi,
> I tried to redefine the features, you can see bellow, but the hypervisor
> still keeps it as active.. According to my understanding anc compute.js it
> should change to shutdown isnt it? Is there any state diagram fot STATE as
> is for LVM_STATE?
>
I've just noticed you are using a ttylinux image. This image does not
support ACPI events, therefore if you want to shutdown the VM you will have
to use the "cancel" action (hard shutdown).
The states diagram [1] represents both LCM_STATE and STATE only when the
STATE is ACTIVE, the LCM_STATE is considered. You can check the existing
states in the VirtualMachine.h file (enum VmState and enum LcmState)
Cheers
[1] http://opennebula.org/_media/documentation:rel3.8:states-simple.png
>
> Milos
>
> BTW current:
> [oneadmin at kvasi occi]$ onevm show 112
> VIRTUAL MACHINE 112 INFORMATION
> ID : 112
> NAME : one-112
>
> USER : oneadmin
> GROUP : oneadmin
> STATE : ACTIVE
> LCM_STATE : SHUTDOWN_POWEROFF
> RESCHED : No
> HOST : kvasi.k13132.local
> START TIME : 06/07 08:02:30
> END TIME : -
> DEPLOY ID : one-112
>
> VIRTUAL MACHINE MONITORING
> NET_TX : 0K
> NET_RX : 0K
>
> USED MEMORY : 0K
> USED CPU : 0
>
> PERMISSIONS
> OWNER : um-
> GROUP : ---
> OTHER : ---
>
> VIRTUAL MACHINE TEMPLATE
> CPU="1"
> DISK=[
> CLONE="NO",
> DATASTORE="emc-spc",
> DATASTORE_ID="104",
> DEV_PREFIX="hd",
> DISK_ID="0",
> DRIVER="raw",
> IMAGE="ttylinux-per",
> IMAGE_ID="76",
> IMAGE_UNAME="oneadmin",
> PERSISTENT="YES",
> READONLY="NO",
> SAVE="YES",
> SOURCE="/dev/vg-c/lv-one-76",
> TARGET="hda",
> TM_MAD="shared_lvm",
> TYPE="FILE" ]
> FEATURES=[
> ACPI="yes" ]
> GRAPHICS=[
> LISTEN="0.0.0.0",
> PORT="6012",
> TYPE="vnc" ]
> MEMORY="1024"
> NAME="one-112"
>
> OS=[
> ARCH="i686",
> BOOT="hd" ]
> RAW=[
> TYPE="kvm" ]
> TEMPLATE_ID="1"
> VCPU="2"
> VMID="112"
>
>
> VIRTUAL MACHINE HISTORY
> SEQ HOST REASON START TIME PROLOG_TIME
> 0 kvasi.k13132.lo user 06/07 08:02:56 0d 00h01m45s 0d 00h00m00s
> 1 kvasi.k13132.lo none 06/07 08:04:56 0d 00h01m14s 0d 00h00m00s
> [oneadmin at kvasi occi]$
> [oneadmin at kvasi occi]$ occi-compute show 112
> <COMPUTE href='http://127.0.0.1:4567/compute/112'>
> <ID>112</ID>
>
> <USER name='oneadmin' href='http://127.0.0.1:4567/user/0'/>
> <GROUP>oneadmin</GROUP>
> <CPU>1</CPU>
> <MEMORY>1024</MEMORY>
> <NAME>one-112</NAME>
>
> <STATE>ACTIVE</STATE>
> <DISK id='0'>
> <STORAGE name='ttylinux-per' href='http://127.0.0.1:4567/storage/76'/>
> <TYPE>FILE</TYPE>
> <TARGET>hda</TARGET>
> </DISK>
> </COMPUTE>
>
>
> Dne 6.6.2013 17:47, Daniel Molina napsal(a):
>
> Hi Miloš,
>
>
> On 6 June 2013 10:37, Miloš Kozák <milos.kozak at lejmr.com> wrote:
>
>> Hi,
>>
>> Template:
>> ACPI="yes"
>> CPU="1"
>> DISK=[
>> IMAGE="ttylinux-per",
>> IMAGE_UNAME="oneadmin" ]
>> GRAPHICS=[
>> LISTEN="0.0.0.0",
>> TYPE="vnc" ]
>> MEMORY="1024"
>> NAME="ttylinux"
>> OS=[
>> ARCH="i686",
>> BOOT="hd" ]
>> RAW=[
>> TYPE="kvm" ]
>> TEMPLATE_ID="1"
>> VCPU="2"
>>
>>
>> States:
>>
>> 110 oneadmin oneadmin one-110 shut 0 0K kvasi.k131 0d
>> 00h01
>> occi-compute show 110
>> <COMPUTE href='http://127.0.0.1:4567/compute/110'>
>> <ID>110</ID>
>> <USER href='http://127.0.0.1:4567/user/0' name='oneadmin'/>
>> <GROUP>oneadmin</GROUP>
>> <CPU>1</CPU>
>> <MEMORY>1024</MEMORY>
>> <NAME>one-110</NAME>
>> <STATE>ACTIVE</STATE>
>> <DISK id='0'>
>> <STORAGE href='http://127.0.0.1:4567/storage/76'
>> name='ttylinux-per'/>
>> <TYPE>FILE</TYPE>
>> <TARGET>hda</TARGET>
>> </DISK>
>> </COMPUTE>
>>
>> After poweroff:
>>
>> onevm show 110
>> VIRTUAL MACHINE 110 INFORMATION
>> ID : 110
>> NAME : one-110
>> USER : oneadmin
>> GROUP : oneadmin
>> STATE : ACTIVE
>> LCM_STATE : SHUTDOWN_POWEROFF
>> RESCHED : No
>> HOST : kvasi.k13132.local
>> START TIME : 06/03 10:00:41
>> END TIME : -
>> DEPLOY ID : one-110
>>
>> VIRTUAL MACHINE MONITORING
>> NET_RX : 0K
>> NET_TX : 0K
>> USED MEMORY : 0K
>> USED CPU : 0
>>
>> PERMISSIONS
>> OWNER : um-
>> GROUP : ---
>> OTHER : ---
>>
>> VIRTUAL MACHINE TEMPLATE
>> ACPI="yes"
>> CPU="1"
>> DISK=[
>> CLONE="NO",
>> DATASTORE="emc-spc",
>> DATASTORE_ID="104",
>> DEV_PREFIX="hd",
>> DISK_ID="0",
>> DRIVER="raw",
>> IMAGE="ttylinux-per",
>> IMAGE_ID="76",
>> IMAGE_UNAME="oneadmin",
>> PERSISTENT="YES",
>> READONLY="NO",
>> SAVE="YES",
>> SOURCE="/dev/vg-c/lv-one-76",
>> TARGET="hda",
>> TM_MAD="shared_lvm",
>> TYPE="FILE" ]
>> GRAPHICS=[
>> LISTEN="0.0.0.0",
>> PORT="6010",
>> TYPE="vnc" ]
>> MEMORY="1024"
>> NAME="one-110"
>> OS=[
>> ARCH="i686",
>> BOOT="hd" ]
>> RAW=[
>> TYPE="kvm" ]
>> TEMPLATE_ID="1"
>> VCPU="2"
>> VMID="110"
>>
>> VIRTUAL MACHINE HISTORY
>> SEQ HOST REASON START TIME
>> PROLOG_TIME
>> 0 kvasi.k13132.lo user 06/03 10:00:56 0d 00h04m10s 0d
>> 00h00m00s
>> 1 kvasi.k13132.lo none 06/03 10:05:26 0d 00h00m26s 0d
>> 00h00m00s
>> [oneadmin at kvasi occi]$ occi-compute show 110
>> <COMPUTE href='http://127.0.0.1:4567/compute/110'>
>> <ID>110</ID>
>> <USER name='oneadmin' href='http://127.0.0.1:4567/user/0'/>
>> <GROUP>oneadmin</GROUP>
>> <CPU>1</CPU>
>> <MEMORY>1024</MEMORY>
>> <NAME>one-110</NAME>
>> <STATE>ACTIVE</STATE>
>> <DISK id='0'>
>> <STORAGE name='ttylinux-per' href='http://127.0.0.1:4567/storage/76'/
>> >
>> <TYPE>FILE</TYPE>
>> <TARGET>hda</TARGET>
>> </DISK>
>> </COMPUTE>
>>
>>
>> Is that all you need to know? BTW it is ONE 3.8.3.
>>
>
>
> The state of the VirtualMachine is ACTIVE, that's why OCCI expose also
> the ACTIVE state
>
> onevm:
>
> STATE : ACTIVE
> LCM_STATE : SHUTDOWN_POWEROFF
>
> occi:
>
> <STATE>ACTIVE</STATE>
>
> The VirtualMachine will stay in that state until it disappears from the
> hypervisor, if the action does not succeed after a while the lcm_state will
> change to running.
>
> I think the problem is how you are specifying the ACPI attr, you have to
> include it in a FEATURES section:
> http://opennebula.org/documentation:rel4.0:kv
> <http://opennebula.org/documentation:rel4.0:kvmg#features>
>
>
>
> mg#features <http://opennebula.org/documentation:rel4.0:kvmg#features>
>
> Cheers
>
>
>>
>> BTW I am sorry for resending. First, I sent it directly outside of the
>> mailing list..
>>
>>
>> Dne 3.6.2013 9:53, Daniel Molina napsal(a):
>>
>> Hi,
>>
>>
>> On 2 June 2013 10:10, Miloš Kozák <milos.kozak at lejmr.com> wrote:
>>
>>> Hi,
>>> thank you for the answer. I tried to verify that. It is quite easy to
>>> sent LCM_STATES to XML, thought. But at this point I would rather tried to
>>> resolve it with VM_STATE. I am afraid that there might be a bug. Source
>>> from compute.js:
>>>
>>> function VMStateBulletStr(vm){
>>> var vm_state = vm.COMPUTE.STATE;
>>> var state_html = "";
>>> switch (vm_state) {
>>> case "INIT":
>>> case "PENDING":
>>> case "HOLD":
>>> case "STOPPED":
>>> case "SUSPENDED":
>>> case "POWEROFF":
>>> state_html = '<img
>>> style="display:inline-block;margin-right:5px;;"
>>> src="images/yellow_bullet.png" alt="'+vm_state+'" title="'+vm_state+'" />';
>>> break;
>>> case "ACTIVE":
>>> case "DONE":
>>> state_html = '<img
>>> style="display:inline-block;margin-right:5px;"
>>> src="images/green_bullet.png" alt="'+vm_state+'" title="'+vm_state+'"/>';
>>> break;
>>> case "FAILED":
>>> state_html = '<img
>>> style="display:inline-block;margin-right:5px;" src="images/red_bullet.png"
>>> alt="'+vm_state+'" title="'+vm_state+'"/>';
>>> break;
>>> };
>>> return state_html;
>>> }
>>>
>>> As I read it, the XML should contain states as poweroff and so on, but
>>> it gives only done, pending, done and active. I ran small script on a VM:
>>>
>>> until [ `sleep 0.7` ]; do occi-compute show 109 | grep STATE; done;
>>>
>>> And triggered all thinkable commands on the VM. When I tryed poweroff
>>> and shutdown it prevailed in ACTIVE. That is why I think there might by a
>>> problem..
>>>
>>> I tried to resolve it on my own, but I dont know ruby....
>>>
>>>
>> Could you check the states with onevm show and confirm that the action
>> (shutdown/power off) doesn't fail. Note that you will need ACPI activated
>> on your VMs to run these actions.
>>
>> Cheers
>>
>>
>>>
>>> Thanks for answer,
>>> Milos
>>>
>>> Dne 26.4.2013 11:23, Daniel Molina napsal(a):
>>>
>>> Hi ,
>>>
>>>
>>> On 25 April 2013 09:28, Miloš Kozák <milos.kozak at lejmr.com> wrote:
>>>
>>>> Hi,
>>>> I am running opennebula 3.8.3 and OCCI self-service portal. My problem
>>>> is that the VM indication is misleading. There 3 statuses - green, yellow,
>>>> red. When I stop VM it turns to yellow, if anything is wrong red.. that is
>>>> perfectly correct but the VM is indicated by green for shutdown, poweroff
>>>> and all other statuses.. I was trying to fix compute.js, but it didnt
>>>> worked out.. So I assume there is a deeper problem? Can you confirm that?
>>>>
>>>
>>> When using OCCI the VM xml that is sent in a OCCI /compute/:id GET
>>> request include the VM_STATE [1].
>>>
>>> VM_STATE=%w{INIT PENDING HOLD ACTIVE STOPPED SUSPENDED DONE FAILED
>>> POWEROFF}
>>>
>>> The problem is that the states you are looking for are LCM_STATES.
>>>
>>> LCM_STATE=%w{LCM_INIT PROLOG BOOT RUNNING MIGRATE SAVE_STOP SAVE_SUSPEND
>>> SAVE_MIGRATE PROLOG_MIGRATE PROLOG_RESUME EPILOG_STOP EPILOG
>>> SHUTDOWN CANCEL FAILURE CLEANUP UNKNOWN HOTPLUG SHUTDOWN_POWEROFF
>>> BOOT_UNKNOWN BOOT_POWEROFF BOOT_SUSPENDED BOOT_STOPPED}
>>>
>>> If you want to include this information you have to modify the
>>> VirtualMachineOCCI class to include these states [2]
>>>
>>> Hope this helps
>>>
>>> [1]
>>> https://github.com/OpenNebula/one/blob/release-3.8.3/src/oca/ruby/OpenNebula/VirtualMachine.rb
>>> [2]
>>> https://github.com/OpenNebula/one/blob/release-3.8.3/src/cloud/occi/lib/VirtualMachineOCCI.rb
>>>
>>>
>>>> Thank you, Milos
>>>> _______________________________________________
>>>> Users mailing list
>>>> Users at lists.opennebula.org
>>>> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>>>>
>>>
>>>
>>>
>>> --
>>> Daniel Molina
>>>
>>>
>>>
>>
>>
>> --
>> Join us at OpenNebulaConf2013 <http://opennebulaconf.com/> in Berlin,
>> 24-26 September, 2013
>> --
>> Daniel Molina
>> Project Engineer
>> OpenNebula - The Open Source Solution for Data Center Virtualization
>> www.OpenNebula.org | dmolina at opennebula.org | @OpenNebula
>>
>>
>>
>> _______________________________________________
>> Users mailing list
>> Users at lists.opennebula.org
>> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>>
>>
>
>
> --
> Join us at OpenNebulaConf2013 <http://opennebulaconf.com/> in Berlin,
> 24-26 September, 2013
> --
> Daniel Molina
> Project Engineer
> OpenNebula - The Open Source Solution for Data Center Virtualization
> www.OpenNebula.org | dmolina at opennebula.org | @OpenNebula
>
>
>
> _______________________________________________
> Users mailing list
> Users at lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>
>
--
Join us at OpenNebulaConf2013 <http://opennebulaconf.com/> in Berlin, 24-26
September, 2013
--
Daniel Molina
Project Engineer
OpenNebula - The Open Source Solution for Data Center Virtualization
www.OpenNebula.org | dmolina at opennebula.org | @OpenNebula
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opennebula.org/pipermail/users-opennebula.org/attachments/20130610/74df8cb0/attachment-0002.htm>
More information about the Users
mailing list