Marco, The ONAP teardown script that I mentioned has now been checked in here: https://gerrit.onap.org/r/#/c/27305/
Note that it currently assumes the username/password of demo/onapdemo to delete the DNS zones. We can refactor the credentials later. Thanks, Gary From: PLATANIA, MARCO (MARCO) [mailto:plata...@research.att.com] Sent: Wednesday, January 03, 2018 9:03 AM To: Yunxia Chen <helen.c...@huawei.com>; jwagant...@linuxfoundation.org; Gary Wu <gary.i...@huawei.com>; FORSYTH, JAMES <jf2...@att.com>; FLOOD, JERRY <jf9...@att.com>; Gildas Lanilis <gildas.lani...@huawei.com>; Yang Xu (Yang, Fixed Network) <yang....@huawei.com>; Kang Xi <kang...@huawei.com> Cc: onap-discuss@lists.onap.org Subject: Integration meeting 01/03 - minutes Hi All, Here are the minutes from the Integration meeting today (01/03). Feel free to integrate in case I missed something. Note I added one item to the list, that is the current Integration meeting conflict with other meetings. In case we want to move it, Wednesday at 9 AM EST or Tuesday at 9 AM or 10 AM EST works best for me. Thanks, Marco 1. Amsterdam Maintenance release testing status APPC is not planning to release any maintenance version, unless some high/highest bug comes up. Dan Timoney mentioned during the last TSC call that he was planning to release a maintenance version for CCSDK and SDNC, but we need to confirm. Jimmy Forsyth is planning to release a maintenance version of AAI. He will work with his Team on properly tagging the maintenance containers such that the Integration Team can pick up the right ones. Marco will reach out to Lusheng (DCAE), Pam (Policy), and Dan Timoney (CCSDK/SDNC – for confirmation) to know their maintenance release planning (if any). 2. Progress and demo for automate the following two use cases a. vFW / vDNS Jerry implemented the vFW closed loop automation in Robot. This however works only 30% of the time of less. It needs to be verified whether there is an issue with the way the ONAP platform is installed. Jerry will investigate further. For vLB, the policy validation step fails, so the entire process exits unsuccessfully. We need Policy Team onboard to verify where the problem is. One issue that many people have seen is that the operational policy takes long time to reload in the vFW use case. This means that it may take 20-30 minutes before Policy kicks in during closed loop. b. vCPE Kang developed custom scripts that can be used to automate the vCPE use case. This doesn’t include the vCPE service design via SDC, which is done manually. However, once the service is created, those scripts can be used to create multiple vCPE instances. Kang has also created a tutorial that people can follow to run the entire vCPE use case. Kang will share the link to the tutorial with the community. Kang also found some instability with VPP-based VNFs in vCPE. He’s working with developers on solving those issues and improving the VNF stability. c. VoLTE Yang is working with the Chinese Teams to extend VFC to include a generic VNF manager that can be used to instantiate generic VNFs. The timeline is Beijing Release, so it will not be available for Amsterdam Maintenance Release testing. Helen is also working with the Chinese Teams on some early automation of the VoLTE use case based on simulations. 3. Beijing release related topics: a. Benchmark Project Benchmark project presented at the Santa Clara event and approved as Integration sub-project. As such, Helen will be the PTL. The focus of the sub-project is performance measurements. We need the Team onboard to better understand to what extent they plan to run benchmarks (E2E? Single components? Etc.) b. Progress on those automation with Jenkins? The ultimate goal is to have Jenkins jobs that install the ONAP platform, run at least the vFW and vLB/vDNS closed loops, and then clean up the environment. Some of this is already available (i.e. ONAP installation), something will be integrated when available (i.e. vFW and vLB/vDNS automation). Gary worked with Wind River folks to create a script that cleans up the resources allocated by OpenStack during ONAP installation. Gary will provide a link to this script such that everyone can use it after deleting an ONAP instance. Although these scripts are in the Integration repository, they run from a Jenkins instance external to LF. They will eventually run in LF Jenkins. Gary opened a ticket against the LF Helpdesk to show up the results of automated testing directly on LF-managed websites. Jessica will escalate this. c. What can we cover at Beijing release? Not so much progress happened. We’ll need to raise this point again, perhaps in the next TSC meeting (tomorrow 01/04). d. Is OOM ready for Integration team to use now? Very critical, Gildas will raise this point during the TSC call tomorrow (01/04). We need to assess the OOM progress. If Heat needs to be fully supported and modified to include HA requirements, we need to find resources and start to work pretty soon. 4. Weekly call The Integration meeting conflicts with OOM and Use-case Subcommittee meetings (at least). Do we keep this time (i.e. Weds at 10 AM EST/ 7 AM PST/ 10 PM China) or do we want to change?
_______________________________________________ onap-discuss mailing list onap-discuss@lists.onap.org https://lists.onap.org/mailman/listinfo/onap-discuss
