Re: TR : [onap-tsc] [IMPORTANT] OOM status update for RC0

Krzysztof Opasiak

Hi Xu Ran,

On 29.03.2021 03:40, Xu Ran wrote:
Dear Opasiak and Catherine,

I have noticed that the block thing of UUI’s merge into OOM is:

They’re all related to the new NLP server, however, since there’re some
several problems appearing in these two commits, we accept the result to
delay this new part to Istanbul release. We’ll abandon these two commits
and make further changes. And the H release plan of UUI will *NOT*
include this part.
No need to abandon. You can always squash the two and move ahead with a
new patchset on one of the reviews;)

As for the commit:

This commit only contains the changes of E2E slicing service which are
some minor ones.
Yes + there is an implementation of NLP server introduced by commit
7643cc5b373b167000d676c48d741e830081f3ab but I expect that as NLP is
delayed to Istanbul then this is just disabled by default?

We hope that OOM team can review this change and merge
it. If there’s any problem with the E2E part, we’ll be free to change.
 And also, Doctor @Dong Wang is the sponsor of the new NLP part, he is
also available to response the problem of this part.

We will attend today’s PTL meeting to make some more explanation, hope
we can reach an agreement.
Gating seems to be fine on those patches. My only concern, that I
expressed in the ticket is what is the rationale for not providing this
container update before the deadline which is M3?

Your REQ tickets say that you don't expect a need to update OOM
configuration for this release...

Xu Ran
China Mobile Research Institute
No.32 Xuanwumen west street,Xicheng District, Beijing 100053, China

Mobile: +86 15010868144
Email: <>

On 03/27/2021 03:05,Krzysztof Opasiak via<>
<> wrote:

Dear Catherine,

On 26.03.2021 19:13, Lefevre, Catherine wrote:

Dear OOM Team,

David and I we are trying to understand what it is left from OOM
that the project team(s) need to consider before the next PTL
call (3/29)

We understand from the TSC call (3/25) - SO, SDNC and UUI had an
but it was not clear that there were other projects.

Here is our understanding and the path to move forward.

After the following items are solved then we need to stabilize the
release and no more accept any OOM code submission except

for *versioning, documentation and show stoppers blocking E2E

#1 UUI

Remaining code submitted to change version is

I have reviewed it and did not notice anything wrong so I
believe it is
ready to merge

The UUI action was related to:

Not the patch that you linked above. The one that you linked has been
created *yesterday* 40 mins before the TSC meeting which is 4 weeks
after the deadline for starting OOM review with new containers for
Honolulu release.

In my personal opinion it's very unfair for other project teams to
accept such changes with a monthly delay. Think how much extra features
could CPS or DCAE Team implement if they had extra month for the

The following items - we will follow-up if these are required
for RC1

It does not look like new code but more modifications of config
to build
UUI Container. To be confirmed by *XU*



This is a *brand new* microservice that they are willing to add to ONAP
It's not a change in config, it's not any kind of bugfix, it's a brand
new Helm chart under UUI project.

It was uploaded to gerrit a little bit after the deadline but it was
first sent to me via email so we decided to give it a chance.
Unfortunately when we start reviewing it more deeply multiple issues
been discovered but never fixed and I believe that we cannot compromise
the quality because of the deadline.

#2 SO

We check the remaining open defects,

SO-3584 -


(can we merge the code)?

As stated in the review, the patch itself is fine (that's why it has +2
from  me and Sylvain) but we are waiting for a successful gating in
patch that you referred below as this is actually where SO will start
making use of that. If we merge it just now it would be just a blind
merge as without the patch above we don't know if it really works.

SO-3590 -


(waiting for review)

Waiting for gating...



- suggestion is to descope the remaining item of SO factoring to
Istanbul - too many code submitted and not yet reviewed.




- can we descope and shift it to Istanbul?



-- (version update + 1 big fix) Dan T to review the comments
from Morgan

I believe that Dan has already fixed the issue reported by Morgan. We
are waiting for gating to confirm this.


--  verified job ok, ready for review

I need to follow up with Alexander on this patch as I missed his
comment. Sorry for that.

CPS - OK For Honolulu


) verified job ok, ready for review

It's marked as WIP as Bruno is working to address comments from
revision. Nevertheless changes would not require releasing a new
container image. They just need to switch to use certInitializer for
their https termination on ingress.

Considering the progress that they made in this release, their nice
cooperation with us almost from the beginning of the release and amount
of comments that they addressed I believe that it's fair to take this
patch even as a bugfix after RC0 as it would affect OOM only

OOF - OK For Honolulu


verified job ok, ready for review

It's not ready, it doesn't work. I got some hint from Krishna how to
it so I'll give it a try this evening if it works we'll merge if not we
just postpone it. It's not a functional change it's just enabling
logging to STDOUT for OOF so I believe that it can be merge even
after RC0

Holmes - What's level of confidence that it will work? Can we
move this
to Istanbul.


Guangrong Fu made recently some fixes but due to bad certs in DMAAP we
haven't had a chance to see it in the action. After we have gating
results we'll be able to say anything more.

AAI - OK For Honolulu (required for Certification)


- verified job ok, ready for review

Waiting for gating results...
Code-wise looks fine. If it passes in the gating I'll merge it.

Multicloud - submitted before RC0


- verified job ok, ready for review

Already reviewed;)
As I'm coauthor of this patch, Sylvain is the one to merge it but he is
waiting for jenkins job review before doing that. Code-wise looks fine
so as soon as we have results from jenkins I hope it can be merged.

*OOM Team*- do we miss  anything else? THANK YOU

There is one more bugfix from the SDNC (CCSDK) team that has been
submitted yesterday:

I'm working with them to find the best solution but I'm sure we'll sort
this out quickly especially that it's just a change in the helm charts
that is easy to gate.

Best regards,
Krzysztof Opasiak
Samsung R&D Institute Poland
Samsung Electronics

Krzysztof Opasiak
Samsung R&D Institute Poland
Samsung Electronics

Join to automatically receive all group messages.