...
- Revise LF infrastructure roadmap to include the timeline and TSC-58 content in advance of TSC meeting Kenny Paul Former user (Deleted)
- Krzysztof Opasiakto provide vulnerability procedure overview next week 2019-02-25
/* Vulnerability Management Procedure is being updated so I'd like to delay this till the procedure is approved by TSC to avoid confusion. Hopefully the overview will be delivered next week. */ - Yang Xu add integration sanity passing in milestone checklist
-
(PTLs): Review ONAP codebase license scan, Feb. 2019 sent by Steve Winslow on 2/18. (reminder)
- 2/11: Migrate Docker image releases from Nexus3 into Docker Hub (Architecture Independence) by Dublin M4.
<2-15-2019> Preparing a preso for review of all known infrastructure change requests for review at PTL meeting 2-25-2019 - Oom values.yaml or integration repo manifest (
TSC-86 - Lock down docker image tag name source of truth - oom values.yaml or integration repo manifest - A: both but manifest is the source SUBMITTED ) Michael O'Brien
if not covered - see TSC 2019-01-10
Quick discussion on nailing down whether we need a yaml override of the deployable
docker image tags in the oom repo (understanding is no) - with the integration docker manifest
(manifest is currently a copy of the oom values.yaml tags - not the reverse)
If not - then we need a documented procedure wiki/RTD on running a derived values.yaml override for the entire system before deployment
AAF is only an example here
https://git.onap.org/oom/tree/kubernetes/aaf/charts/aaf-service/values.yaml?h=casablanca#n28
image: onap/aaf/aaf_service:2.1.8
drives
onap/aaf/aaf_service,2.1.8
<1-7/2019> Integration team to document the procedure.<01/13/2019> Answer from Integration team: docker manifest under integration repo is the source of truth, and is used by Integration team to override OOM values.yaml when deploying ONAP in Openlab. See the instructions at the bottom of page https://onap.readthedocs.io/en/casablanca/submodules/integration.git/docs/onap-oom-heat.html#onap-oom-heat-template
<1-21/2019> Procedure should include Timers, dependencies, etc.
just need to define all the --set and -f overrides like the following
sudo helm deploy onap local/onap --namespace $ENVIRON -f $DISABLE_CHARTS_YAML -f $DEV0_YAML $APPENDABLE_ENABLED_FLAGS --verbose
<1-28-2019>: Document how to handle the latest 3rd party image from github.
<2019-02-19>: One over-ride that will be shared, parent over-ride for docker versions, Michael O'Brien to modify TSC-86 to reflect process. Still looking at tool to institute changes for all projects (single commit). Mike Elliott FREEMAN, BRIAN D Gary Wu Sylvain Desbureaux to follow up offline.
It should also consider the problem identified with the Casablanca Maintenance Release:https://lists.onap.org/g/onap-tsc/message/4641 - Helm Chart Transfer: 4 projects have been identified for trial: A&AI, Log, Policy, APPC, CLAMP and ONAP CLI. Mike Elliott will work with PTL to gather/transfer knowledge and create training materials. Training to be done the week of Jan 14 and PTL to provide their commitment or not by M1 on Jan 24.<1-7/2019>: Work in progress. Schedule might shift to another week. Michael will follow-up with Mike. Training will be open to everybody. The objective is that each team will take over the management of their Helm Charts.
<1-28/2019>: Meeting is organized on 1/29 with LF. Training will be organized early in February. Additional information on the next PTL call (2/4)<2019-2-19: Meeting on this topic today - review the process - develop training material. OOM subcharts complete - A&AI this week> - What is the plan for the June ONAP Developers Forum - dates/location/etc Kenny Paul Former user (Deleted)
Zoom Chat Log
| Anchor | ||||
|---|---|---|---|---|
|
06:02:49 From Jimmy Forsyth (AT&T) : https://wiki.onap.org/pages/viewpage.action?pageId=8231933#MeetingHelp,Requests,Changes,Hosting&Recording-RecordingPrivileges:Recordings
06:04:19 From Kenny Paul (LFN) : Thanks Jimmy :-)
06:05:59 From Pamela Dragosh : I think M2 non-root is simply a question.
06:06:08 From Catherine Lefevre : correct
06:06:16 From Catherine Lefevre : assessment at M2
06:09:07 From Kenny Paul (LFN) : Pam asks a question about duration between M2 & M3
06:11:24 From Kenny Paul (LFN) : Catherine: What is represented is a 3 week duration which was approved by the TSC - anything not complete by M4 will be descoped
06:11:41 From Michael O'Brien : Mike - logging on
06:11:59 From Kenny Paul (LFN) : #topic Risks
06:12:50 From Catherine Lefevre : M4 - April 4, 2019. This date shifted when M1 was shifting from Jan 17th to Jan 24th
06:15:33 From Kenny Paul (LFN) : No portal representative on the call
06:26:51 From Kenny Paul (LFN) : Geo-redundancy - need additial lab resources
06:36:20 From Eric Debeau : Orange is not sure to get all physical resources for inetgration
06:38:14 From Catherine Lefevre : ok thanks Eric
06:38:39 From Catherine Lefevre : I will follow-up about T-Lab
06:39:08 From Catherine Lefevre : we need to understand issues per SB-xx from Yang;; e-mail sent to yang to collect issues per lab
06:41:28 From Kenny Paul (LFN) : Seshu- concerned about upgrade - Need to ensure that there is coordination (Regarding #15)
06:44:00 From Kenny Paul (LFN) : #17 OOM- Helm charts lower risk now - Bengt working w/ Mike
06:44:47 From Kenny Paul (LFN) : Footprint optimization (alpine vs ubantu) meeting this week
06:46:27 From Kenny Paul (LFN) : #18 Pam - SDC stretch goal but at least need SDC to answer questions
06:54:30 From Eric Debeau : #17 OOM- Helm charts lower risk now - Bengt working w/ Mike Footprint optimization (alpine vs ubantu) meeting this week: when is scheduled the meeting ?
06:54:36 From Pamela Dragosh : What does OVP stand for?
06:55:52 From Brian : Is LF driving the Optimize all operations or are some of them really ONAP projects run by th ecommunity ?
06:56:30 From Sai Seshu : Opnfv verification program
06:56:41 From Pamela Dragosh : Thanks Seshu!
06:57:11 From Sai Seshu : Anytime Pam :)
06:59:18 From Catherine Lefevre : action #Kenny, jessica - review original requests from PTLs captured under TSC-58 and the LF plan presented today
07:04:06 From Michael O'Brien(LOG, Amdocs) : YUL - I hear it is those Canadian DC's - the one in Montreal
07:04:19 From Kenny Paul (LFN) : #action @jess, @kenny @jim provide time frames for infrascructure plans
07:05:19 From Kenny Paul (LFN) : #topic seccom - java
07:05:35 From Michael O'Brien(LOG, Amdocs) : +1 on 11 (I am skipping 9, 10) - some pom changes at least
07:05:36 From Kenny Paul (LFN) : want to update to java v11
07:05:50 From Pamela Dragosh : + 1 Policy already planned to move to v11
07:05:55 From Kenny Paul (LFN) : impact on projects
07:06:19 From Kenny Paul (LFN) : ?
07:06:22 From Michael O'Brien(LOG, Amdocs) : for El-alto
07:06:40 From Pamela Dragosh : As long as not required, we need to assess if there are any problems.
07:06:42 From Vijay VK (AT&T) : -1 for DCAE for Dublin
07:07:13 From Michael O'Brien(LOG, Amdocs) : +1 on Java 11 el-alto only - due to unknown impacts beyond config changes - need full integration testing
07:07:26 From Catherine Lefevre : @Pawel - keep this as part of your El-Alto request
07:07:39 From Catherine Lefevre : we have not yet colleced El-Alto requirements but it will come soon
07:08:16 From Michael O'Brien(LOG, Amdocs) : Right, there is also some investigation on JAXB and JPA support needed
07:08:54 From Michael O'Brien(LOG, Amdocs) : first step could be compile = 11, but source compatability is 8 - (as in no 11 code changes)
07:08:55 From Catherine Lefevre : ODL upgrade to java 8.0 no alignment before Frankfurt
07:09:00 From Brian : ODL moves to java 11 support in Sodium
07:09:04 From Catherine Lefevre : so APPC/CDS/CCSDK impacted
07:09:12 From Michael O'Brien(LOG, Amdocs) : k8s is still 1.11.5 - looking towards 1.12+ all the way to 1.13 - el-alto lately
07:09:17 From Catherine Lefevre : sorry brian - i miscaptued the java version
07:10:35 From Jonathan Gathman : When I say “Client” it also applies to running infrastructure like ODL. You CAN compile your code to lower Java Source, and still make sure it all runs in JDK 11… However, that is definitely late to mess with in Dublin.
07:16:10 From Kenny Paul (LFN) : #topic don't break the build
07:16:41 From Michael O'Brien(LOG, Amdocs) : +1 on docker manifest only for release - as different devs run different sets of docker images based on use/non-use of the manifest
07:17:13 From Kenny Paul (LFN) : Yang reviewed the slides attached to the meeting minutes
07:18:11 From Michael O'Brien(LOG, Amdocs) : Ideally the CD system has a record of deployment logs - correlated to the gerrit reviews submitted - so we can track issues from (n) previous short/long regression cD runs