Re: [onap-tsc] 答复: Re: 答复: Re: Proposal for Holmes project
Hi Junjie, I just quickly review the Holmes proposal and DMaaP proposal. I have some confusions, please help me to clarify. Thanks. For the scope part, does the filtering function will be implemented in DMaaP project? Or What is the difference between Holmes filtering and data filtering in DMaaP? I guess there are some overlap. In addition, the compression and aggregation function of Holmes is quite different with the function in DMaaP. Holmes is in the service level, and DMaaP is in the data level. Do I understand correctly? Because you use the same words, it is a little confusion for me. Regards, Yuan 刘媛 / Liu Yuan 网络技术研究所 / Department of Network Technology 中国移动通信研究院 / China Mobile Research Institute Mobile: +86 15810024078 Email: liuyuan...@chinamobile.com 发件人： liu.junj...@zte.com.cn 发送时间： 2017-05-18 10:02 收件人： JI; l...@research.att.com 抄送： onap-tsc; fu.guangrong 主题： [onap-tsc]答复: Re: 答复: Re: Proposal for Holmes project Hi As my understanding, the rules in Holmes just only define how to correlate the enents, the policy consist of three parts, the events for triger policy, the rules of policy and the action of policy. So i think Holmes is belong to DCAE which mostly related to analysis. For align project, we think now DCAE in ONAP provide the framework, Holmes project can be a sub-project as one DCAE special application project, we image there will be more and more DCAE application projects in more areas in DCAE. considering easy to manage DCAE project in the future , we wish DCAE project as framework and the 'Holmes' as a application project. Project risk division can be good for first release. 刘军杰 liujunjie GA产品经理 GA Product Manager 服务及MANO产品部/系统产品 Service & MANO Product Dept./System Product 原始邮件 发件人： ＜l...@research.att.com＞; 收件人：刘军杰10037236; 抄送人： ＜onap-tsc@lists.onap.org＞;付光荣10144542; ＜ma...@research.att.com＞; ＜c...@research.att.com＞; ＜pdrag...@research.att.com＞; ＜rshac...@research.att.com＞; 日 期 ：2017年05月18日 03:45 主 题 ：Re: [onap-tsc] 答复: Re: Proposal for Holmes project Junjie, I still have not received any feedback from ZTE regarding the deep dive of the Holmes project. We need to decide how to align projects. Thanks, Lusheng AT&T Labs ONAP DCAE From: ＜onap-tsc-boun...@lists.onap.org＞ on behalf of "JI, LUSHENG (LUSHENG)" ＜l...@research.att.com＞ Date: Friday, May 12, 2017 at 12:47 AM To: "liu.junj...@zte.com.cn" ＜liu.junj...@zte.com.cn＞, "gilb...@mail.linuxfoundation.org" ＜gilb...@mail.linuxfoundation.org＞, "GILBERT, MAZIN E (MAZIN E)" ＜ma...@research.att.com＞ Cc: "onap-tsc@lists.onap.org" ＜onap-tsc@lists.onap.org＞, "fu.guangr...@zte.com.cn" ＜fu.guangr...@zte.com.cn＞ Subject: Re: [onap-tsc] 答复: Re: Proposal for Holmes project ***Security Advisory: This Message Originated Outside of AT&T *** Reference http://cso.att.com/EmailSecurity/IDSP.html for more information. Mazin and Junjie, As I was preparing the DCAE project proposal, I had the same impression as Mazin that Holmes correlation engine appeared to align well with DCAE's micro service analytics architecture. Hence I put in a placeholder for possible harmonization with Holmes. However this afternoon I had some time to briefly scan through Holmes Mercury source code and now I am actually feeling less confident about this impression. Holmes implementation appears to consist of code around two major functional components, a DB (mysql) and a rule engine (Drools). The rules are stored in DB, and loaded into Drools engine. The Drools engine monitors Active MQ topics for alarms, then matches alarms with some rules. Is this high level understanding correct? If so, this actually appears to resemble some of the Policy Engine implementations more than DCAE analytics (cc-ing Pam, who is Policy Engine lead). I think we need to schedule a deep dive meeting finding how to best align. Given now it is already Friday afternoon in China, shall we target early next week? Thanks, Lusheng Original message From: liu.junj...@zte.com.cn Date: 5/11/17 11:10 PM (GMT-05:00) To: gilb...@mail.linuxfoundation.org, "GILBERT, MAZIN E (MAZIN E)" ＜ma...@research.att.com＞ Cc: fu.guangr...@zte.com.cn, onap-tsc@lists.onap.org Subject: [onap-tsc] 答复: Re: Proposal for Holmes project Hi Yes, I agree with your suggestion, Holmes will be a Microservice application running in DCAE just like all analytics Microservices. Holmes is not a independence application without DCAE structure. 刘军杰 liujunjie GA产品经理 GA Product Manager 服务及MANO产品部/系统产品 Service & MANO Product Dept./System Product 原始邮件 发件人： ＜ma...@research.att.com＞; 收件人： ＜dr6...@att.com＞;刘军杰10037236; 抄送人： ＜onap-tsc@lists.onap.org＞;付光荣10144542; 日 期 ：2017年05月12日 10:40 主 题 ：Re: [onap-tsc] Proposal for Holmes project Liu, Interesting. I need to spend time looking at Holmes a bit cl
[onap-tsc] 答复: Re: 答复: Re: Proposal for Holmes project
Hi As my understanding, the rules in Holmes just only define how to correlate the enents, the policy consist of three parts, the events for triger policy, the rules of policy and the action of policy. So i think Holmes is belong to DCAE which mostly related to analysis. For align project, we think now DCAE in ONAP provide the framework, Holmes project can be a sub-project as one DCAE special application project, we image there will be more and more DCAE application projects in more areas in DCAE. considering easy to manage DCAE project in the future , we wish DCAE project as framework and the 'Holmes' as a application project. Project risk division can be good for first release. 刘军杰 liujunjie GA产品经理 GA Product Manager 服务及MANO产品部/系统产品 Service & MANO Product Dept./System Product 原始邮件 发件人： ＜l...@research.att.com＞ 收件人：刘军杰10037236 抄送人： ＜onap-tsc@lists.onap.org＞付光荣10144542 ＜ma...@research.att.com＞ ＜c...@research.att.com＞ ＜pdrag...@research.att.com＞ ＜rshac...@research.att.com＞ 日 期 ：2017年05月18日 03:45 主 题 ：Re: [onap-tsc] 答复: Re: Proposal for Holmes project Junjie, I still have not received any feedback from ZTE regarding the deep dive of the Holmes project. We need to decide how to align projects. Thanks, Lusheng AT&T Labs ONAP DCAE From: ＜onap-tsc-boun...@lists.onap.org＞ on behalf of "JI, LUSHENG (LUSHENG)" ＜l...@research.att.com＞ Date: Friday, May 12, 2017 at 12:47 AM To: "liu.junj...@zte.com.cn" ＜liu.junj...@zte.com.cn＞, "gilb...@mail.linuxfoundation.org" ＜gilb...@mail.linuxfoundation.org＞, "GILBERT, MAZIN E (MAZIN E)" ＜ma...@research.att.com＞ Cc: "onap-tsc@lists.onap.org" ＜onap-tsc@lists.onap.org＞, "fu.guangr...@zte.com.cn" ＜fu.guangr...@zte.com.cn＞ Subject: Re: [onap-tsc] 答复: Re: Proposal for Holmes project ***Security Advisory: This Message Originated Outside of AT&T *** Reference http://cso.att.com/EmailSecurity/IDSP.html for more information. Mazin and Junjie, As I was preparing the DCAE project proposal, I had the same impression as Mazin that Holmes correlation engine appeared to align well with DCAE's micro service analytics architecture. Hence I put in a placeholder for possible harmonization with Holmes. However this afternoon I had some time to briefly scan through Holmes Mercury source code and now I am actually feeling less confident about this impression. Holmes implementation appears to consist of code around two major functional components, a DB (mysql) and a rule engine (Drools). The rules are stored in DB, and loaded into Drools engine. The Drools engine monitors Active MQ topics for alarms, then matches alarms with some rules. Is this high level understanding correct? If so, this actually appears to resemble some of the Policy Engine implementations more than DCAE analytics (cc-ing Pam, who is Policy Engine lead). I think we need to schedule a deep dive meeting finding how to best align. Given now it is already Friday afternoon in China, shall we target early next week? Thanks, Lusheng Original message From: liu.junj...@zte.com.cn Date: 5/11/17 11:10 PM (GMT-05:00) To: gilb...@mail.linuxfoundation.org, "GILBERT, MAZIN E (MAZIN E)" ＜ma...@research.att.com＞ Cc: fu.guangr...@zte.com.cn, onap-tsc@lists.onap.org Subject: [onap-tsc] 答复: Re: Proposal for Holmes project Hi Yes, I agree with your suggestion, Holmes will be a Microservice application running in DCAE just like all analytics Microservices. Holmes is not a independence application without DCAE structure. 刘军杰 liujunjie GA产品经理 GA Product Manager 服务及MANO产品部/系统产品 Service & MANO Product Dept./System Product 原始邮件 发件人： ＜ma...@research.att.com＞ 收件人： ＜dr6...@att.com＞刘军杰10037236 抄送人： ＜onap-tsc@lists.onap.org＞付光荣10144542 日 期 ：2017年05月12日 10:40 主 题 ：Re: [onap-tsc] Proposal for Holmes project Liu, Interesting. I need to spend time looking at Holmes a bit closer. From a first glance, the rule designer seems like a GUI for writing policies. The correlation engine is an algorithm that ingest data and provides signatures based on those correlations. If my interpretation is right, then I suggest putting the correlation engine as a Microservice running in DCAE just like all analytics Microservices. The GUI is similar to CLAMP. CLAMP allows you to design open and closed loop systems besides just putting the policies and rules. CLAMP connects with ONAP during both design and runtime. Something to think about. thanks mazin On May 11, 2017, at 10:29 AM, ROSE, DANIEL V ＜dr6...@att.com＞ wrote: ***Security Advisory: This Message Originated Outside of AT&T *** Reference http://cso.att.com/EmailSecurity/IDSP.html for more information. The dcae proposal includes holm
[onap-tsc] 答复: Re: 答复: Re: Proposal for Holmes project
Hi Holmes implementation appears to consist of code around two major functional components, a DB (mysql) and a rule engine (Drools). The rules are stored in DB, and loaded into Drools engine. The Drools engine monitors Active MQ topics for alarms, then matches alarms with some rules. Is this high level understanding correct? Response： Yes, your understanding is correct. Given now it is already Friday afternoon in China, shall we target early next week? Response：OK, we can discuss these issues in next week early. 刘军杰 liujunjie GA产品经理 GA Product Manager 服务及MANO产品部/系统产品 Service & MANO Product Dept./System Product 原始邮件 发件人： ＜l...@research.att.com＞ 收件人：刘军杰10037236 ＜gilb...@mail.linuxfoundation.org＞ ＜ma...@research.att.com＞ 抄送人：付光荣10144542 ＜onap-tsc@lists.onap.org＞ ＜pdrag...@research.att.com＞ 日 期 ：2017年05月12日 12:48 主 题 ：Re: [onap-tsc] 答复: Re: Proposal for Holmes project Mazin and Junjie, As I was preparing the DCAE project proposal, I had the same impression as Mazin that Holmes correlation engine appeared to align well with DCAE's micro service analytics architecture. Hence I put in a placeholder for possible harmonization with Holmes. However this afternoon I had some time to briefly scan through Holmes Mercury source code and now I am actually feeling less confident about this impression. Holmes implementation appears to consist of code around two major functional components, a DB (mysql) and a rule engine (Drools). The rules are stored in DB, and loaded into Drools engine. The Drools engine monitors Active MQ topics for alarms, then matches alarms with some rules. Is this high level understanding correct? If so, this actually appears to resemble some of the Policy Engine implementations more than DCAE analytics (cc-ing Pam, who is Policy Engine lead). I think we need to schedule a deep dive meeting finding how to best align. Given now it is already Friday afternoon in China, shall we target early next week? Thanks, Lusheng Original message From: liu.junj...@zte.com.cn Date: 5/11/17 11:10 PM (GMT-05:00) To: gilb...@mail.linuxfoundation.org, "GILBERT, MAZIN E (MAZIN E)" ＜ma...@research.att.com＞ Cc: fu.guangr...@zte.com.cn, onap-tsc@lists.onap.org Subject: [onap-tsc] 答复: Re: Proposal for Holmes project Hi Yes, I agree with your suggestion, Holmes will be a Microservice application running in DCAE just like all analytics Microservices. Holmes is not a independence application without DCAE structure. 刘军杰 liujunjie GA产品经理 GA Product Manager 服务及MANO产品部/系统产品 Service & MANO Product Dept./System Product 发件人： ＜ma...@research.att.com＞ 收件人： ＜dr6...@att.com＞刘军杰10037236 抄送人： ＜onap-tsc@lists.onap.org＞付光荣10144542 日 期 ：2017年05月12日 10:40 主 题 ：Re: [onap-tsc] Proposal for Holmes project Liu, Interesting. I need to spend time looking at Holmes a bit closer. From a first glance, the rule designer seems like a GUI for writing policies. The correlation engine is an algorithm that ingest data and provides signatures based on those correlations. If my interpretation is right, then I suggest putting the correlation engine as a Microservice running in DCAE just like all analytics Microservices. The GUI is similar to CLAMP. CLAMP allows you to design open and closed loop systems besides just putting the policies and rules. CLAMP connects with ONAP during both design and runtime. Something to think about. thanks mazin On May 11, 2017, at 10:29 AM, ROSE, DANIEL V ＜dr6...@att.com＞ wrote: ***Security Advisory: This Message Originated Outside of AT&T *** Reference http://cso.att.com/EmailSecurity/IDSP.html for more information. The dcae proposal includes holmes as a sub part (・ Holmes: this is an OpenO project that performs similar/related role to DCAE. Further information is needed for identifying path for normalization.) Did you talk to anyone from dcae about overlap? If so what was the differences between your projects? Just trying to wrap my head around the difference for my own edification. Thanks, Daniel Rose ECOMP / ONAP com.att.ecomp 732-420-7308 From: onap-tsc-boun...@lists.onap.org [mailto:onap-tsc-boun...@lists.onap.org] On Behalf Of liu.junj...@zte.com..cn Sent: Thursday, May 11, 2017 4:32 AM To: onap-tsc@lists.onap.org Cc: fu.guangr...@zte.com.cn Subject: [onap-tsc] Proposal for Holmes project Dear all Holmes project provides alarm correlation and analysis for Telecom cloud infrastructure and services, including host, vim, vnf and ns. Homels aims to find the reason which cause the fail or degradation of services by digging into the ocean of events collected from different levels of Telecom cloud. Holmes is a application that processes events published by managed resources or other applications that detect specific conditions. Based on
