[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: 答复:Re: [tosca] RE: Comments about workflow //Re:RE: [tosca] Re: Try to submit a comment//Re:[tosca] Your Public Review for TOSCA Simple Profile in YAML V1.1 CSPRD02 has been announced
Hi Huabing, Meeting starts in 1 hour and 10 minutes. Luc From:
"zhao.huabing@zte.com.cn" <zhao.huabing@zte.com.cn> Hi Matt,
原始邮件
发件人:
MattRutkowski; 收件人:赵化冰10201488;
抄送:luc.boutier@fastconnect.fr; Paul.Lipton@ca.com; tosca@lists.oasis-open.org;
日期: 2017-02-21 04:44:36
主题:Re: [tosca] RE: Comments about workflow //Re:RE: [tosca] Re: Try to submit a comment//Re:[tosca]
Your Public Review for TOSCA Simple Profile in YAML V1.1 CSPRD02 has been announced
Hi Huabing,
Thank you for your quick response. I totally agree that no language has been ‘specified’ in the TOSCA xml 1.0. Actually I think it's appropriate to keep the Workflow modeling approach open in TOSCA. And the leverage of existing language like BPMN or BPEL can definitely enhanced the portability
and interoperability rather than weakened them. Below is extracted from TOSCA xml 1.0, which clearly explains the portability and interoperability benefits by leveraging existing standards of workflow languages. Plans defined in a Service Template describe the management aspects of service instances, especially their creation
and termination. These plans are defined as process models, i.e. a workflow of one or more steps. Instead of providing another language for defining process models, the specification relies on existing languages like BPMN or BPEL.
Relying on existing standards in this space facilitates portability and interoperability, but any language for defining process models can be used.
The TOSCA metamodel provides containers to either refer to a process model (via
Plan Model Reference) or to include the actual model in the plan (via Plan Model). A process model can contain tasks (using BPMN terminology) that refer to operations of Interfaces of Node Templates (or operations defined by the Node Types specified
in the typeattribute of the Node Templates, respectively), operations of Interfaces of Relationship Templates
(or operations defined by the Relationship Types specified in the typeattribute of the Relationship Templates,
respectively), or any other interface (e.g. the invocation of an external service for licensing); in doing so, a plan can directly manipulate nodes of the topology of a service or interact with external systems. My suggestion is that TOSCA can evolve in a compatible and open manner, and it should continuously support the standard Workflow modeling languages such as BPMN or BPEL. It might be helpful to introduce a new approach for workflow modeling inside TOSCA in
some cases , however, it should not be exclusive. Could we add this topic to this week's TOSCA Simple Profile WG meeting so we can have deep discussions? Thanks, Huabing Original Mail
Hi Huabing, As I have been quite involved in the workflow works I will share my opinion and explain the motivations around that. First thing is that no language has been ‘specified’ really in the TOSCA xml 1.0 and the suggestion didn’t really provide any strong answer. There were also no specific definitions on how to write them
in a portable manner and in regard to the actual elements that exists in TOSCA (nodes, operations etc.). The goal of having workflows/plans in the TOSCA spec is to allow definition of workflows that suits to the TOSCA need, the TOSCA model, and that would
allow portability. Moreover, the Simple profile in YAML has a strong target of making TOSCA simple to write in yaml and using a yaml description for imperative workflows was very logical in this perspective. It also allowed
to write a workflow logic that really suits the TOSCA needs with a very clear description of what an orchestrator should support so he can manage portable TOSCA templates. Finally, as I said the recommendation from XML spec being just a ‘non-portable recommendation’ I don’t think that the fact that your tool and some others made their choice with probably their own decision
on how to model TOSCA specific things (outside the TC) can be a valid argument for pushing it. I also know some other TOSCA tools that are using the TOSCA 1.1 definitions and believe this is a more portable implementation in regard of TOSCA. That said, I guess that it is easy to translate the standard and portable workflow parsing into your specific implementation and eventually if you have been working with some other projects make that an
opensource contribution. I suppose that tools that have been parsing the TOSCA xml specification have anyway to work on supporting the YAML specification now. My 2 cents, Luc From: <tosca@lists.oasis-open.org> on behalf of Paul Lipton <Paul.Lipton@ca.com> Thank you for adding this discussion as JIRA issue TOSCA-318. FYI, Luc Boutier made a tremendous contribution to the workflow aspects of Simple Profile in YAML v1.1, so you may find his opinion
to be helpful. That said; I encourage interested TC members to voice their opinions as comments to this issue.
Regards, Paul
From: zhao.huabing@zte.com.cn [mailto:zhao.huabing@zte.com.cn]
Hi Paul and Chet, Many thanks for the useful information and guidelines. So I guess I can add my two cents here. I notice that the draft of "TOSCA Simple Profile in YAML Version 1.1" is trying to define a new workflow DSL inside TOSCA instead of using existing standards such as BPMN/BPEL which is recommended in the Topology and Orchestration Specification for Cloud
Applications Version 1.0[1]. Below is the description of plan(workflow) in the V1.0 Spec: planLanguage: This attribute denotes the process modeling language (or metamodel) used to specify the plan. For example, “http://www.omg.org/spec/BPMN/20100524/MODEL”
would specify that BPMN 2.0 has been used to model the plan. TOSCA does not specify a separate metamodel for defining plans. Instead, it is assumed that a process modelling language (a.k.a. metamodel) like BPEL [BPEL 2.0] or BPMN [BPMN 2.0] is used to define plans. The specification favours
the use of BPMN for modelling plans. Since many open source and property orchestration implementation have already adopted TOSCA and used the BPMN/BPEL as the process modelling language, this incompatible change may force them to refactor their implementation to adapt to the new Spec, wich
will definitely cost a lot of time and work. As far as I know, OPEN-O[2], Open-TOSCA[3] and some of the internal projects of ZTE will encounter this problem. So I'm wondering could we keep standard modelling language, in particular, such as BPMN and BPEL, as an option in the next version of TOSCA Spec?
What do you think? [1]
http://docs.oasis-open.org/tosca/TOSCA/v1.0/TOSCA-v1.0.html [3]http://www.iaas.uni-stuttgart.de/OpenTOSCA/ Original Mail
Sender: <Paul.Lipton@ca.com>; To: <chet.ensign@oasis-open.org>;zhaohuabing10201488; CC: <tosca@lists.oasis-open.org>;
<member-services@oasis-open.org>;mengzhaoxing10024238; Date: 2017/02/16 23:02 Subject: RE: [tosca] Re: Try to submit a comment//Re:[tosca] Your Public Review for TOSCA Simple Profile in YAML V1.1 CSPRD02 has been announced Hi Huabing, I am Co-Chair of the TOSCA TC. Chet is correct that TC Members can always post to our mailing list, and also create issues in our issue tracking system, based on JIRA. Additionally, I will
send you an orientation email shortly, which will explain our organizations and the other ways that you can get involved.. If you have any questions, please don’t hesitate to contact me or my esteemed co-chair, John Crandall, directly. Regards, Paul From:
tosca@lists.oasis-open.org[mailto:tosca@lists.oasis-open.org]
On Behalf Of Chet Ensign Hello Huabing, Welcome to OASIS and the TOSCA TC! I am delighted to read about open-o. I love to see how projects are finding value in the spec.
To answer your question, yes - as a member of the TC, you should send your comments to the main TOSCA TC mailing list. The tosca-comment@ mailing list is for use by those who are outside the TC.
Your comments on the spec should be tracked and addressed just the same as any coming to the comment@ mailing list.
Please let me know how I can help as you engage with the work here.. Best regards, /chet On Wed, Feb 15, 2017 at 11:51 PM, <zhao.huabing@zte.com.cn>wrote: I'm Huabing from ZTE(http://www.zte.com.cn),
a new member of OASIS and TOSCA TC. ZTE is adopting TOSCA as the modelling approach in its NFV products, ZTE is also a premier member of OPEN-O(https://www.open-o.org),
which is an open source orchestration project targeted on the world’s most important and innovative telecommunications and cable operators. In OPEN-O, ZTE contributed the model designer which leverage TOSCA for the NFV, SDN and End to End service modelling.
That's why we're so interested in the TOSCA TC and join it. I noticed the public review announcement for TOSCA Simple Profile in YAML Version 1.1 sent out earlier today. I try to follow the below instructions in the Public Review announcement to submit a comment, but the "Send A Comment" button at the TC public
home page and the link direct me to a non-member feedback instruction page(I already logged in). Comments may be submitted to the TC by any person through the use of the OASIS TC Comment Facility which can be accessed via the button labeled "Send A Comment" at the top of the TC public home page, or directly at: http://www.oasis-open.org/committees/comments/form.php?wg_abbrev=tosca So, as a member, am I supposed to post the comments directly to
tosca@lists.oasis-open.org? It's my pleasure to join TOSCA TC and get this precious opportunity to work with you all. Thanks, Huabing Original Mail
Sender: <chet.ensign@oasis-open.org>; To: <tosca@lists.oasis-open.org>; Date: 2017/02/16 00:42 Subject: [tosca] Your Public Review for TOSCA Simple Profile in YAML V1.1 CSPRD02 has been announced Members of the TOSCA TC, Your 15 day public review for Your Public Review for TOSCA Simple Profile in YAML V1.1 CSPRD02 has been announced. The review ends March 2nd. You can find the announcement at
https://www.oasis-open.org/news/announcements/15-day-public-review-for-tosca-simple-profile-in-yaml-version-1-1-ends-march-2nd.
Unfortunately, your notification email to
cmwg-chair@dmtf.orgbounced. Please keep in mind the OASIS requirements for handling comments [1]. Non-TC member feedback can only be submitted to the TC's comment list
tosca-comment@lists.oasis-open.org. The TC must have someone subscribed to this mail list to monitor comments. All submitted comments must be acknowledged by the TC. In addition, the TC
needs to maintain a log of comments received and their resolutions. The comment resolution log will need to be avaiable when you begin your next public review. A simple comment resolution log template is available in OpenDocument [2] and Office [3] format.
Let me know if you have any questions regarding the review or next steps. === Additional references: [1]
https://www.oasis-open.org/resources/tcadmin/handling-the-comments-received-during-a-public-review
--
|
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]