Attachments (including meeting recordingAttachments (including meeting recording):
Attendees:
| Status | OPEN | IN PROGRESS | ON HOLD | DONE | CANCELLED |
|---|
START RECORDING
| 1 | SONAR coverage report |
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
Graphgraph not included in OOM deployment. How can we make it an essential part of the ONAP deployment.
Step 0: Make it stable. Fix the docker build Jenkins jobs for graphgraph, may need to upgrade to a version of npm to be acceptable to graphgraph. Needs to fit into the LF global-job jobs.
How do we accelerate migration using A&AI and the RunTime DB solution
Proposed a topic to be explored by AI&I and other ONAP WGs.
High impact, common interest among most CSPs
Status Quo:
Most operators have multiple, siloed, possibly overlapping topology and inventory (domain specific) databases
Ultimate goal is to integrate ALL CSP inventory, and Topology databases into a Unified AI&I DB for ALL VNFs (VMs/Containers), PNFs, plus future inventory constructs/technologies.
May take considerable time/investment (both are short).
Targeted Objectives
- Fast integration/consolidation time
- Minimize impact to business systems during migration
- Reduce throw away code
- Reduce throw away integration work
- Improve Service/Resource LCM & SA, especially for 5G
How to accelerate the migration to Unified Database?
- Phased Externalization of key PNF/VNF/CNF attributes currently stored in multiple inventory and topology silos.
- Simplify the eventual consolidation of Inventory and Topology DBs using proposed automation tools and advanced discovery capabilities supporting the above
- Leverage AI&I and possibly the ONAP Runtime Config DB and other ONAP WGs to help solve the problem
- Proposed topic can be explored by the AI&I and other ONAP WGs.
- Need to expand this preliminary list
https://sonarcloud.io/organizations/onap/projects
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
James Forsyth, the main reason behind these discrepancies is the fact that SonarQube is based on version 6.7.1 and SonarCoud is version 8+
After migrating to SonarCloud, we fetch updates they do automatically without the need of doing any manual upgrade. Since we migrated, few things have changed. For example:
- LF does no longer mark warnings in the reports, only pass or fail
- LF is no longer able to manage individual quality profiles per language. Whatever latest support is there, we fetch this automatically in our side.
- Jacoco plugin has changed quite a bit, we need to make sure our reports are now xml based instead of exec. I think this can be the main reason why you are facing coverage drops. Can you please check? This is what Sonar mentioned to me in detail:
You can find more info in the wiki :https://wiki.onap.org/display/DW/SonarCloud+migration+from+SonarQube
| Jira Legacy | ||||||
|---|---|---|---|---|---|---|
|
Graphgraph doesn't parse the latest AAI schema correctly. Need someone to investigate
| Jira Legacy | ||||||
|---|---|---|---|---|---|---|
|
| Jira Legacy | ||||||
|---|---|---|---|---|---|---|
|
2 Spring beans with the same name is disallowed in Spring boot 2
Sparky is using the AAF certificate interaction with Portal but a self-signed certificate to talk to search-data-service. The search-data-service key story was removed.
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
The Intel/Wind River lab was relocated into a new data center last week. The only change required to connect to the new location is to update the VPN IP in the .ovpn file. Here is the change:
[sgooch@atl-sgooch-lx ~]$ diff -u pod-onap-01.ovpn.org pod-onap-01.ovpn
--- pod-onap-01.ovpn.org 2020-01-13 13:03:03.960856010 -0500
+++ pod-onap-01.ovpn 2020-01-09 18:27:03.560549576 -0500
@@ -5,7 +5,7 @@
client
dev tun
proto tcp
-remote 192.55.48.241 443
+remote 146.152.204.133 443
resolv-retry infinite
auth-nocache
nobind
NNI Use Case - discovering 2 topologies from the network
Use case is asking for the ability in AAI to link an underlay and overlay together by having a cousin relationship between 2 PNFs. Asking Chandra Cinthalato verify if that edge rule can be added for the Frankfurt release.
Per Chandra: "I think pnf -> pnf edge is redundant and not needed as we can traverse (thru a logical or physical interface) from pnf -> p-interface/l-interface -> pnf."
- shashikanth.vh will let the AAI team know if that solves the problem
Chandra Cinthala please note: According to shashikanth.vh, the team requires a relationship from an overlay topology node to an underlay topology node; by using the p-interface in overlay we cannot traverse to underlay node. Suggest to allow an edge for PNF→PNF, which is consistent with RFC 8345 model
Merged and released yesterday; issue is closed.
Jira Legacy server System Jira columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution serverId 4733707d-2057-3a0f-ae5e-4fd8aff50176 key AAI-2617
Tracking for Frankfurt:
Jira Legacy server System Jira columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution serverId 4733707d-2057-3a0f-ae5e-4fd8aff50176 key AAI-2677
Search guard container no longer available so elasticsearch-sg won't build. Search guard allows us to secure the elastic communication, which in older versions was not available but newer versions have support for secure communication.
https://www.elastic.co/blog/dear-search-guard-users
Consider opendistro as replacement for search-guard
Can we consider this one:
https://docs.search-guard.com/latest/search-guard-community-edition
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
Francis Paquette need evaluation of feasibility in Frankfurt for sparky
Graph Graph Images are missing from nexus and it needs to be resolved in order for elalto deployment to be successful.
OOM-2161
How did this docker image get pushed to nexus and why was it removed from nexus?
Graphgraph was removed from the default helm chart, but we still need to get these jobs fixed.
- James Forsythwill look at it and ask Former user (Deleted)for guidance on this - looks like some jjbs are missing and configs are probably wrong, too.
- Ondrej F will reach out to Pavel and see if we can figure out why the build and stage jobs are failing.
Assigned
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
| Jira Legacy | ||||||
|---|---|---|---|---|---|---|
|
SDC wants to move up to Cassandra 3 in the shared db environment.
| Jira Legacy | ||||||
|---|---|---|---|---|---|---|
|
https://old-docs.janusgraph.org/0.3.2/version-compat.html
Venkata Harish Kajur tried schema service, a couple of dependency changes were required to get it to start. Eclipselink that is currently configured is incompatible.
<groupId>javax.activation</groupId>
<artifactId>activation</artifactId>
is also incompatible
Needs more investigation for other repos
Need new common image (which Alpine image would we pick?)
- James Forsythwill ask cl664y@att.comabout the status of a committer who has left the company (Tian Lee has left Amdocs, do we revoke committer status?)
Peng He is a potential committer on ESR (currently only Jimmy can commit ESR code)
Goal for Frankfurt is 55%
AAI-Common 58.7
AAI-Champ 58.3
AAI-DataRouter 58
AAI-GraphAdmin 56.6
Schema Service 56.7
Sparky 57.3
Spike 56.3
- Steve Blimkie will check with architects around the utility of champ and spike in the context of ongoing support in ONAP
- James Forsyth will set them to read-only (gizmo, champ, spike) and El Alto will be the final release version of those microservices
Jira Legacy server System Jira serverId 4733707d-2057-3a0f-ae5e-4fd8aff50176 key AAI-2732 - James Forsyth will remove the jenkins jobs on those repos
| Jira Legacy | ||||||
|---|---|---|---|---|---|---|
|
sparky-fe Jira Legacy server System Jira serverId 4733707d-2057-3a0f-ae5e-4fd8aff50176 key AAI-2711
graphgraph Jira Legacy server System Jira columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution serverId 4733707d-2057-3a0f-ae5e-4fd8aff50176 key AAI-2713
Release 6 (Frankfurt) proposed use cases and functional requirements
Use Case Realization Call: September 25, 2019
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
We have created REQ-158 -
for this use case. For AAI impacts, we have createdAAI-2600 - A&AI support of Network Slicing Demo in Frankfurt.Reg. details of the AAI impacts, this is still being worked out, as there is an alternative proposal, and we are trying to align w.r.to scope & impacts. We will share more details in the coming days. The overall use case will also be presented to the Architecture Sub-committee for approval (based on the f2f discussions that happened in Antwerp 2 weeks ago).View file name AAI Weekly - 2019-12-04.mp4 height 250
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
- James Forsyth No status yet, will ask on UCR call today
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
We are currently reviewing the list of items we found in the BBWF BBS testing and once we sort through them we will communicate the results back to ONAP community.
Regarding AAI specifically, we do not foresee any new requirements for Frankfurt, but we would like to conclude the review first before jumping to final conclusions.
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
- James Forsyth No status yet, will ask on UCR call today. Followed up, no representation from use case. Will look for response via email.
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
We are not expecting any changes at this time. Tech Mahindra is working on populating a&ai with the VNF specific information from K8S. There is a chance that we realize some changes to the schema, but we don’t expect any changes to the code.
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
During run-time solution there are AAI impacts. Run-time solution is not planned for Frankfurt release. Only Import Service Catalog solution in SDC is targeted for F release.
AAI Impact highlights:
After Service Order decomposition in SO, CFS Service Instance details will be stored in AAI.
After response from 3^rd^ Party Domain, RFS instance details will be updated in AAI.
The sequence diagram 
depicts the run time view of Third Party Order Activation using the on-boarded service definition and highlights AAI impacts.
ReUse in AAI:
We plan to leverage the SPPartner model, created for CCVPN, to store the 3^rd^ Party details in AAI.
org.onap.aai.domain.yang.SpPartner present in aai-schema v15
- James Forsyth Confirm on UCR call that this is object re-use and test only support from AAI
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
Have done some prototyping in Dublin to ingest topology from external controllers and show them in A&AI. Will need more work to act on notifications from controllers and keep the topology in sync
Jira Legacy server System Jira columns key,summary,type,created,updated,due,assignee,reporter,priority,status,resolution serverId 4733707d-2057-3a0f-ae5e-4fd8aff50176 key AAI-2678
For the 5G Bulk PM use case in R6, I’m attaching a few slides showing how we think AAI can be used as part of the UC realization. We are currently making the following assumptions:
- PNF and VNF models contain status attribute that can indicate when a new instance has been fully instantiated (may still require update to SO workflow)
- AAI will send topology event on AAI-EVENT topic when status attributes are updated
Can you review the slides and the assumptions above and let us know if you see any issue. We can also setup a separate meeting if needed.
Note: I will share the same slides on the SO weekly meeting today.
| View file | ||||
|---|---|---|---|---|
|
Frankfurt Release Requirements
- James Forsyth Follow up on URC call today
What is the purpose of MigrateCloudRegionUpgradeCycle migrator in onap graphadmin migration framework?
- Need William LaMont to check on this by
From Casablanca to Dublin, should these be run?
- v14/MigrateMissingFqdnOnPservers
- v13/MigrateVnfcModelInvariantId
- v13/MigrateModelVer
Migration will check if it's already been run, and won't run again.
Jira task created to track this: AAI-2689
Questions regarding how to use the api
Sparky Running on Node Port: 30220
Status on
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
Pain, pain, pain. 218 items that must be documented currently across the AAI repos, expanding all the time.
1) We need a repo strategy
2) We need a maintenance strategy
3) We need a versioning strategy
4) We need responsible parties for each repo / component
5) Strategy for oparent dependency management
Need to know if the ZTE team who built ESR intends to continue to support this sub-project in Frankfurt. Multiple Jira issues were not worked in Dublin / El Alto, which leads me to think that the ESR project might need to be deprecated. Inviting Bo Lv and Former user (Deleted)to see what the current status is.
Hi, we didn't have resources for ESR in El Alto, and I(Huabing Zhao) and Lv Bo won't be able to contribute to ESR in F release, but Peng He will continue supporting ESR in F release on behalf of ZTE team. Please add him in the loop for the ESR related issues.
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
A kubernetes cluster is installed in WR using RKE
NFS node is 10.12.5.10
Followed these instructions https://docs.onap.org/en/latest/submodules/oom.git/docs/oom_setup_kubernetes_rancher.html?
Need to determine what gets installed, testing opportunities, CSIT, distribution flows, etc.
Since the CCVPN SOTN proposal is in progress already, could we have the v17 schema introduced into master branch to receive the updates for Frankfurt release?
| Jira Legacy | ||||||
|---|---|---|---|---|---|---|
|
See also AAI-CCVPN-Extension-SOTN Proposals for Frankfurt Release and
| Jira Legacy | ||||||
|---|---|---|---|---|---|---|
|
| Jira Legacy | ||||||
|---|---|---|---|---|---|---|
|
| Jira Legacy | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
|
docker pull nexus3.onap.org:10003/onap/babel:1.5.0 - 1.5.1 pushed, will be release version
docker pull nexus3.onap.org:10003/onap/aai-champ:1.5.1 - 1.5-STAGING-latest available, no tagged images
docker pull nexus3.onap.org:10003/onap/aai-gizmo:1.5.0 - 1.5.2 pushed, will be release version
docker pull nexus3.onap.org:10003/onap/aai-data-router:1.5.0 - 1.5.1, will be release version
docker pull nexus3.onap.org:10003/onap/aai-search-data-service:1.5.0 - 1.5-STAGING-latest available, no tagged images
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
docker pull nexus3.onap.org:10003/onap/elasticsearch-sg:1.5.0 - same as search data service
docker pull nexus3.onap.org:10003/onap/aai-model-loader:1.5.0 - 1.5.1 will be release version
docker pull nexus3.onap.org:10003/onap/aai-spike:1.5.0 - 1.5.1 will be release version
docker pull nexus3.onap.org:10003/onap/aai-sparky-be:1.5.0 - 1.5-STAGING-latest, no tagged images
Want to migrate to an actively support maven docker plugin since Spotify docker plugin. Consider fabric.io - an example is aai/resources
Jenkins jobs needed to create / push the common alpine image that the AAI containers will use, still need this completed before we merge the outstanding commits.
| Jira Legacy | ||||||
|---|---|---|---|---|---|---|
|
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
Consider adding glow root profiling for AAI Microservices
FREEMAN, BRIAN D reported that when there are ~2400 vservers returned in a closed loop custom query, AAI takes minutes to respond. We've discussed paging previously, any other ways to optimize this?
PUT 'https://aai.onap:8443/aai/v16/query?format=resource'
{"query":"query/closed-loop","start":"/cloud-infrastructure/cloud-regions/cloud-region/CloudOwner/RegionOne/tenants/tenant/28481f6939614cfd83e6767a0e039bcc/vservers/vserver/6ad9eb6b-d7fc-46d6-8617-4c9b46e7308b"}
| Jira Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|
|
aai-common:1.5.2 (not released yet) will be the el-alto library for springboot 1.5.x microservices. There needs to be some urgency to migrate the microservices to spring boot 2, since 1.5.3+ will not be backward compatible and so dependent applications might not have security updates if they stay on 1.5.2. 1.6.0 (Frankfort, currently master branch) aai-common will be springboot 2 so microservices must be updated by Frankfurt. Springboot 1.5.21 is the latest (and last) version in the 1.5.x spring boot train.
The AT&T team has done an exciting POC at a sprint-a-thon event that they would like to share with the community.
ATT wants to contribute additional UI views to ONAP, want to discuss path forward
...