[one-users] [one-user] AppFlow 4.2 Scalling Function
Carlos Martín Sánchez
cmartin at opennebula.org
Thu Sep 5 02:40:42 PDT 2013
Hi,
>From the documentation [1]:
When an attribute is found, it will take the average value for all the
> running VMs that contain that attribute in the Role.
> If none of the VMs contain the attribute, the expression will evaluate to
> false.
Note that the average is only made from the values of running VMs, any VM
on BOOT, POWEROFF, UNKNOWN, etc. will not be considered for the expression
evaluation.
Regards
PS: Please reply to the mailing list
[1]
http://opennebula.org/documentation:rel4.2:appflow_elasticity#auto-scaling_based_on_metrics
--
Join us at OpenNebulaConf2013 <http://opennebulaconf.com> in Berlin, 24-26
September, 2013
--
Carlos Martín, MSc
Project Engineer
OpenNebula - The Open-source Solution for Data Center Virtualization
www.OpenNebula.org | cmartin at opennebula.org |
@OpenNebula<http://twitter.com/opennebula><cmartin at opennebula.org>
On Thu, Sep 5, 2013 at 5:30 AM, Naza <naza1715 at gmail.com> wrote:
> Thanks Carlos,
>
> I have one further question. If my service role already scale up how the
> elasticity policies work? Is it based on the first vm or from the average
> of all vm.
>
> Let say my elasticity policies as below:
>
> DISK_USED_PERC_SDA1 > 50
>
> Is the value of DISK_USED_PERC_SDA1 only calculated from the first VM only
> or from the average of all vm in the same role?
>
> Regards,
> Naza
>
>
> On Wed, Sep 4, 2013 at 5:54 PM, Carlos Martín Sánchez <
> cmartin at opennebula.org> wrote:
>
>> Hi,
>>
>> What is happening is that the wizard is creating a template with these
>> values:
>> # Periods = 60
>> Period = 3
>>
>> This is a bug [1] that we solved after the final release, but somehow we
>> forgot to include it in the known issues page. I'll add a warning in the
>> documentation with the workaround:
>>
>> $ sudo wget
>> http://dev.opennebula.org/projects/opennebula/repository/revisions/5c7b370f6be5f4f9167d508bdd31b580f05dafd4/raw/src/sunstone/public/js/plugins/oneflow-templates.js-O /usr/lib/one/sunstone/public/js/plugins/oneflow-templates.js
>>
>> $ sudo wget
>> http://dev.opennebula.org/projects/opennebula/repository/revisions/970c34464a1d5a71f3a9dfdd3d77d62883400857/raw/src/sunstone/public/js/plugins/oneflow-services.js-O /usr/lib/one/sunstone/public/js/plugins/oneflow-services.js
>>
>>
>>
>> Best regards
>>
>> [1] http://dev.opennebula.org/issues/2249
>> --
>> Join us at OpenNebulaConf2013 <http://opennebulaconf.com> in Berlin,
>> 24-26 September, 2013
>> --
>> Carlos Martín, MSc
>> Project Engineer
>> OpenNebula - The Open-source Solution for Data Center Virtualization
>> www.OpenNebula.org | cmartin at opennebula.org | @OpenNebula<http://twitter.com/opennebula><cmartin at opennebula.org>
>>
>>
>> On Wed, Sep 4, 2013 at 10:52 AM, Naza <naza1715 at gmail.com> wrote:
>>
>>> In the opennebula sunstone oneflow service elasticity policies, it says
>>> that
>>>
>>> # Periods = Number of periods that the expression must be true before
>>> the elasticity is triggered
>>>
>>> Period = Duration, in seconds, of each period in '# Periods'
>>>
>>> My configuration as follow (refer to the attachment)
>>>
>>> # Periods = 3
>>> Period = 60
>>>
>>> I instantiate this template.
>>>
>>> When I run 'oneflow show' command it shows:
>>>
>>> nebula at z2-hn01:~$ oneflow show 4
>>> SERVICE 4
>>> INFORMATION
>>> ID : 4
>>> NAME : test-oneflow-service
>>> USER : nebula
>>> GROUP : oneadmin
>>> STRATEGY : straight
>>> SERVICE STATE : RUNNING
>>> SHUTDOWN : shutdown
>>>
>>> PERMISSIONS
>>>
>>> OWNER : um-
>>> GROUP : ---
>>> OTHER : ---
>>>
>>> ROLE frontend
>>> ROLE STATE : RUNNING
>>> VM TEMPLATE : 3
>>> CARNIDALITY : 1
>>> MIN VMS : 1
>>> MAX VMS : 1
>>> COOLDOWN : 15s
>>> SHUTDOWN : shutdown
>>> NODES INFORMATION
>>> VM_ID NAME STAT UCPU UMEM
>>> HOST TIME
>>> 21 frontend_0_(service_4) runn 0 8G z2-nc06
>>> 0d 01h01
>>>
>>> ROLE db_master
>>> ROLE STATE : RUNNING
>>> PARENTS : frontend
>>> VM TEMPLATE : 4
>>> CARNIDALITY : 2
>>> MIN VMS : 1
>>> MAX VMS : 3
>>> COOLDOWN : 15s
>>> SHUTDOWN : shutdown
>>> NODES INFORMATION
>>> VM_ID NAME STAT UCPU UMEM
>>> HOST TIME
>>> 22 db_master_0_(service_4) runn 0 8G z2-nc05
>>> 0d 00h59
>>> 23 db_master_1_(service_4) runn 0 8G z2-nc04
>>> 0d 00h22
>>>
>>> ELASTICITY RULES
>>>
>>> ADJUST EXPRESSION EVALS
>>> PERIOD COOL
>>> + 1 MEM_USED_PERC[2.8] > 1.5 37 /
>>> 3s 60s
>>> + 1 DISK_USED_PERC_SDA2[--] > 50 0 / 6
>>> 3s 60s
>>>
>>>
>>> LOG
>>> MESSAGES
>>> 09/04/13 15:26 [I] New state: DEPLOYING
>>> 09/04/13 15:31 [I] New state: RUNNING
>>> 09/04/13 16:05 [I] Role db_master scaling up from 1 to 2 nodes
>>> 09/04/13 16:05 [I] New state: SCALING
>>> 09/04/13 16:07 [I] New state: COOLDOWN
>>> 09/04/13 16:08 [I] New state: RUNNING
>>>
>>> Questions:
>>> 1) EVALS value is truncated when it has to digit number (37)
>>> 2) I set 'Period = 60' but it shows 3s
>>> 3) I set '# Periods = 3' but it shows 6
>>> 4) Refer to the LOG, it took about half an hour to scale up. I am not
>>> sure why it is not after (# Periods * Period) which is about 3 minutes (3 *
>>> 60s).
>>>
>>> Please correct me if I misunderstand the oneflow settings.
>>>
>>> Regards,
>>> Naza
>>>
>>>
>>> _______________________________________________
>>> Users mailing list
>>> Users at lists.opennebula.org
>>> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opennebula.org/pipermail/users-opennebula.org/attachments/20130905/b4c34bda/attachment-0002.htm>
More information about the Users
mailing list