[chrisdutz]

Description

Problem

On Linux and macOS, the IT framework never detects ports that are already in use. EnvUtils.searchAvailablePorts() then hands out ports another process listens on, and a node fails to start:

StartupException: Failed to start [Config Node service], because [IoTDB: failed to start Config Node service,
because Could not create ServerSocket on address /127.0.0.1:16490.]
java.net.BindException: Address already in use

The test then fails in setUp with "After 30 times retry, the cluster status check failed: ... Empty configNode set". On my machine this hit two ITs in one SimpleIT/TableSimpleIT run, because another local process was listening on 127.0.0.1:16490.

Cause

listPortOccupationUnix() parses lsof -iTCP -sTCP:LISTEN,TIME_WAIT -P -n and takes the local address from column 9. A line splits into 10 columns, and column 9 is the state:

java 1234 user 77u IPv6 0x1a2b 0t0 TCP 127.0.0.1:6667 (LISTEN)
0    1    2    3   4    5      6   7   8              9

"(LISTEN)".endsWith(":" + port) is never true, so the method always returns an empty map. This also disables the PORT_OCCUPIED handling in AbstractEnv.handleProcessStatus(), which uses the same method to find and restart a node that holds a port.

Fix

Read the address from column 8. The Windows variant (netstat -aon -p tcp, address in column 1) was already correct. A Javadoc line shows the column layout.

Verification

• The layout is the same on both platforms:

  • macOS 27, lsof 4.91: ... TCP 127.0.0.1:16490 (LISTEN), 10 columns.
  • Ubuntu 26.04, lsof 4.99.4: ... TCP 127.0.0.1:6667 (LISTEN) and ... TCP [::1]:6668 (LISTEN), 10 columns, address in column 8.

• Called EnvUtils.listPortOccupationUnix(...) directly on macOS, on master and with this change:

  Port	Before	After
  16490 (held by another process)	{}	{16490=<pid>}
  15292 (held by another process)	{}	{15292=<pid>}
  16491 (free)	{}	{}

• spotless:check passes for integration-test.

• IT runs with this change, on macOS: EdgeIT (7 tests) and IoTDBWithoutNullAllFilterIT (SimpleIT, 11 tests, 5 skipped) pass.

---

This PR has:

• been self-reviewed.
• added comments explaining the "why" and the intent of the code wherever would not be obvious for an unfamiliar reader.

---

Key changed/added classes (or packages if there are too many classes) in this PR

• integration-test/src/main/java/org/apache/iotdb/it/env/cluster/EnvUtils.java

🤖 Generated with Claude Code

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 48.05%. Comparing base (8a5e5ca) to head (8627538).

Additional details and impacted files

@@             Coverage Diff              @@
##             master   #19070      +/-   ##
============================================
- Coverage     48.05%   48.05%   -0.01%     
  Complexity      604      604              
============================================
  Files          5572     5572              
  Lines        410195   410195              
  Branches      54112    54112              
============================================
- Hits         197124   197112      -12     
- Misses       213071   213083      +12     

☔ 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.