Update dependency karafka-web to v0.11.4

[renovate]

This PR contains the following updates:

Package	Change	Age	Confidence
karafka-web (source, changelog)	0.11.3 -> 0.11.4

---

Release Notes

karafka/karafka-web (karafka-web)

v0.11.4

Compare Source

• [Enhancement] Show placeholder rows for partitions with no data during rebalances in health view. The UI now displays all topic partitions (0 to N-1) with "No data available" indicators for partitions currently being rebalanced, preventing confusion from disappearing partitions. Consumer reports now include partitions_cnt field extracted from librdkafka statistics. Consumer schema version bumped to 1.5.0 (breaking change).
• [Enhancement] Track and report UI errors originating from Roda/Puma web processes directly to Kafka errors topic for visibility and debugging. UI errors are dispatched asynchronously from web processes using a dedicated listener.
• [Enhancement] Require Karafka 2.5.2 at minimum and migrate from string-based execution mode comparisons to the new ExecutionMode object API.
• [Enhancement] Increase Web UI processing consumer backoff time to 30 seconds when encountering incompatible schema errors to prevent error spam during rolling upgrades.
• [Enhancement] Add unique id field to error reports to track duplicate error occurrences. Error schema version bumped to 1.2.0 while maintaining backward compatibility with older error formats (1.0.0, 1.1.0) in the Web UI.
• [Enhancement] Add container-aware metrics collection for Docker/Kubernetes environments. The Web UI now reports accurate container memory limits from cgroups (v1 and v2) instead of misleading host metrics, while maintaining full backward compatibility with non-containerized deployments.
• [Enhancement] Add per-message report migration system to handle schema evolution for consumer reports. This allows transparent migration of old report formats (e.g., schema 1.2.x using process[:name]) to current expectations (schema 1.3.0+ using process[:id]), ensuring backward compatibility with reports from older karafka-web versions (≤ v0.8.2) that may still exist in Kafka topics.
• [Change] Reduce max_messages for consumer reports processing from 1000 to 200 to prevent excessive memory usage in large-scale deployments. Processing 1000 messages at once can impact memory consumption significantly in big systems, while 200 messages provides better memory efficiency with negligible impact on throughput.
• [Refactor] Extract metrics collection logic from monolithic Sampler into focused, single-responsibility classes (Metrics::Base, Metrics::Os, Metrics::Container, Metrics::Network, Metrics::Server, Metrics::Jobs) and consumer groups enrichment into dedicated enricher (Enrichers::Base, Enrichers::ConsumerGroups) for improved maintainability and testability.
• [Testing] Add Docker-based integration tests for container metrics collection. Tests verify cgroup v1/v2 detection, memory limit reading, and fallback behavior across multiple containerized scenarios with different resource constraints.
• [Fix] Fix "OS memory used" metric on Linux reporting same value as RSS instead of system-wide memory usage. The metric now correctly sums memory usage across all processes (or all container processes when running in Docker/Kubernetes) to match macOS behavior and original design intent.
• [Fix] Fix crash when processing old consumer reports from schema versions < 1.3.0 (karafka-web ≤ v0.8.2) that used process[:name] field instead of process[:id]. The error undefined method 'to_sym' for nil would occur when these old reports were encountered during upgrades. Reports are now automatically migrated in-place to the current schema format.

---

Configuration

📅 Schedule: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined).

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.