2019-11-05 Security Subcommittee Meeting Notes

Owned by Paweł Pawlak
Last updated: Nov 07, 2019
4 min read

Please see the Minutes of Meetings and recording for the  SECCOM meeting that was held on 5th of November 2019.

Jira No

Summary

Description

Status

Solution

Jira No

Summary

Description

Status

Solution

 

OJSIs summary for El Alto release:

Krzysztof summarized projects and their attitude towards OJSIs security tickets handling:

Still 38 OJSI tickets related to HTTP open while we expose only ~20 HTTP ports. We can close almost half as soon as we get the commit hash-id

Worst performing projects

  • CLI

    • 4 OJSI tickets open

    • 1 ticket with CVE

    • No activity at all

  • Logging

    • 9 OJSI tickets open

    • 1 ticekt with CVE

    • Very limited activity

  • MSB

    • 11 OJSI tickets open

    • 1 ticekt with CVE

    • No activity at all

Could be improved

  • APPC

  • DMAAP

  • INT

  • SDNC

  • SO

Please follow them

  • CLAMP

    • Average response time ~1 day

    • More advanced fixes provided in less than a month

  • Policy

    • Prompt response

    • Following the procedure

    • No open issues.

  • Portal

    • Recognition for the hard work

    • Not everything fixed yet but good progress made

 

 

 

ONAP security maturity assessment

Discussion held at the last PTL call yesterday

PTLs claim that are missing qualified security experts

Idea of SECCOM badging provided per project and per release – discussion point

SECCOM is perceived as group of people pushing PTLs and community to do some security related stuff while it should be the other way around: PTLs are asking SECCOM how they could improve their security.

SECCOM is not about project management and motivating people to do the security.

We should introduce security badging or levels for ONAP projects and sit down together and define what are the requirements for each and every elevel, present those requirements to the projects and at the end of each release to perform the asessment and publish on the release of the project page the list of the projects with their current security status.

KPIs defined with release security maturity should be used.

CII Badging combines multiple areas, including security. To ny proposed cathegorization of CII Badging questions in the security domain to see what would be the scoring for projects within identified cathegories.

As Natacha has stated, CII Badging is not exhaustive list of all security requirements. Krzysztof mentioned that several topics are not covered within CII Badging – like is project exposing anything else that Restful API?

And most of the projects are exposing  DBs outside their cluster or access restriction on all interfaces, or usage of AAF for users management.

So Krzysztof confirmed that from his perspective CII Badging is a starting point and we should define 3-4 security levels. Amy suggested that we should target to projects on what they need to concentrate on. Between release manager, SECCOM and TSC we should decide which requirements are specifically important for the release. Krzysztof shared a point of view that SECCOM presents requirement to TSC and PTLs and then PTLs  are trying to fulfill the requirements, when they believe that they are ready to achieve an another level of security requirement for SECCOM badge, they just come to the SECCOM meeting and present what they have done and how they fulfill it for the new level. Then SECCOM checks that and decide whether badge can be delivered. It was proposed and agreed that we document this proposal and come back to PTL and TSC meetings for a feedback and approval. This change is targeted for the G release.

 

It was agreed to continue discussion in 2 weeks time frame. SECCOM members are requested to provide their feedback.

 

Increased tests code coverage for future ONAP releases

Pierre's proposal:

-define, for each project, the core parts that need intensive testing (an API sensitive project might prioritize API testing so that all are covered, hardened, so that said APIs are robust)

-concentrate coverage on those areas, even if coverage is lower on other parts, so critical parts are better covered and tested.

This might improve OJSI resolution (and/or reduce findings), and better focuses the effort on testing based on available resources.

This would be even better if we could tell Sonar which parts of the code are the critical ones

 

It was agreed to continue discussion in 2 weeks time frame. SECCOM members are requested to provide their feedback.

 

Synch meeting with Architecture Subcommittee

The meeting is scheduled today (5th of November at 4 PM UTC+1). Scope for the discussion:

-Ingress controller (Krzysztof)

-Security architecture document (Hampus)

-ISTIO/AAF discussion (Hampus/Srini)

 

As Natacha is not available, we will cover communication matrix topic for the next synch meeting.

 

Vulnerability management update

Still waiting for inputs from some projects. Conclusion: in order to decrease the size of the vulnerabiliity tables, the easiest way is to keep the latest versions of components.

 

 

 

Frankfurt SECCOM requirements

Most of the projects already updated their descriptions

 

To be further confirmed if all descriptions are present.

 

Nexus-IQ vs. Whitesource benchmarking

Strange results for vulnerabilities – to be consulted with Renan – still waiting for his feedback and Dan’s assessment completion for effective/ineffective

 

 

 

ONAP customized ODL compilation

ODL – Dan contacted Luis for customized package – for both El Alto and Frankfurt releases ONAP is using OpenDaylight Neon SR1.  So, Dan would like the distribution that ODL provides for ONAP custmomized version based on the Neon SR1.

 

 

 

SECCOM meetings timeline

AAF and Architecture Subcommittee timing dependencies to be taken into account, some SECCOM members prefer this modified slot (due to daylight savings change)