server: skip IPv4 member CIDR for IPv6-only security group members

[wido]

Description

Found while reviewing #14037. When a security group rule references another security group, both SecurityGroupManagerImpl and SecurityGroupManagerImpl2 append /32 to the member's IPv4 address without checking it. For a member with an IPv6-only NIC this produces the CIDR null/32.

The KVM agent drops the invalid entry while parsing, but the management server should not generate it. Skip the IPv4 entry when the member has no IPv4 address.

Based on 4.20 on top of #14037.

Types of changes

• Bug fix (non-breaking change which fixes an issue)

Feature/Enhancement Scale or Bug Severity

Bug Severity

• Trivial

How Has This Been Tested?

Unit tests for both implementations with an IPv6-only member, added to SecurityGroupManagerImplTest together with the two tests from #14037. They pass with the fix and fail without it.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 16.41%. Comparing base (376c1b4) to head (93879d0).

Additional details and impacted files

@@             Coverage Diff              @@
##               4.20   #14325      +/-   ##
============================================
- Coverage     16.41%   16.41%   -0.01%     
+ Complexity    13657    13656       -1     
============================================
  Files          5669     5669              
  Lines        501628   501627       -1     
  Branches      60942    60944       +2     
============================================
- Hits          82336    82320      -16     
- Misses       410064   410081      +17     
+ Partials       9228     9226       -2     

Flag	Coverage Δ
uitests	4.16% <ø> (ø)
unittests	17.27% <100.00%> (-0.01%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[wido]

@DaanHoogland can you review? Small follow-up to #14037.

[weizhouapache]

#14037 has been just merged into 4.20.
Could you rebase this with 4.20 too ? @wido

[wido]

#14037 has been just merged into 4.20. Could you rebase this with 4.20 too ? @wido

Done! I thought this one was better for main, but 4.20 is fine with me as well

[weizhouapache]

#14037 has been just merged into 4.20. Could you rebase this with 4.20 too ? @wido

Done! I thought this one was better for main, but 4.20 is fine with me as well

thanks @wido

I just noticed #14037 added two new test classes, any chance to move the new tests in #14037 and this PR into existing SecurityGroupManagerImplTest.java and SecurityGroupManagerImpl2Test.java ?

[weizhouapache]

code lgtm

[DaanHoogland]

code looks good, but agree with the move of the tests @weizhouapache proposes

[wido]

@weizhouapache done, all four tests are now in SecurityGroupManagerImplTest and the four standalone classes are gone. I did not use SecurityGroupManagerImpl2Test because it needs a live MySQL and is excluded in server/pom.xml, so tests in there never run. SecurityGroupManagerImplTest covers both implementations.