completed. Just need to add to release note. Need to make sure to get feedback from stakeholders later
Gerard Nugent investigating ODL to see if we need update for Java 17
Nightly performance Graphs still unstable. Actions agreed:
Daniel Hanrahan to choose and display new 'representative' graphs at top and other data from all test for comparison. Preferably 'extreme' or last volume tests.
(improved version) Now Modules are correctly copied when using an existing ModuleSetTag too Sourabh wil improve testware and stub later to use modules names that are more obviously different (instead of the same name followed by a different sequence number)
Jira
server
ONAP Jira
serverId
425b2b0a-557c-3c0c-b515-579789cceedb
key
CPS-2000
has been raised to address the limitation of s instance based cache for upgrade scenarios. (for cached CM data only, does not affect Ericsson use-cases)
all good, one minor issue with validation of query: The TrustLevel enum values are validated and the trustlevel condition. But the Condition Parameter name is simply ignored and can be anything! This was not considered an issue that need to be fixed by anyone on the call...
documentation of the new query will needed to be add. Halil Cakal will add a ticket for this:
Demo failed first because of incorrect testdata (accidently reverted last fix). Then Sourabh was bale to fix this and demo successfully including upgrade scenario!
Stakeholders expressed whish for DMI-Inventory endpoint to query ModuleSetTags. If really needed this can be done upon request and with relevant priority but was not in scope of originals requirement.
For performance (tetsing) we agreed 10 diff moduleSetTags should be used, Stubs need to be updated to allow this with also different sets of modules being simulated (allowing 90% overlap)
CPS Team hopes to deliver both Epics before the end of next week.
Enable/Disable notifications Anchors using a Post request
Instead of creating a new table Toine Siebelink proposed to modify the exist table with Anchor details to have a new (text) field to store information about delta notifications (which can change in the future)
rajesh kumar to update the proposal accordingly and raise 2 Open issues that needs to be discussed with the architects:
Is it OK to modify Postgres tables instead of using a yang modelled data approach... (this will require Liquibase steps)
generic text field OK for future refinement, maybe consider json?
Sourabh wil improve testware and stub later to use modules names that are more obviously different (instead of the same name followed by a different sequence number)