[APM] Show accurate metrics for containerized applications (#76768)

* adding cgroup fields

* calculate the memory based on the cgroup fields otherwise use the system fields

* updating script

* adding api tests

* using cgroup fields on service map

* addressing PR comment

* fixing unit test

* fixing test

* changing api tests to use snapshot

* removing inactive_files from calculation

* fixing test

* refactoring painless script

* addressing PR comment

* addressing pr comments

Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>

Author: cauemarcondes

x-pack/plugins/apm/common/__snapshots__/elasticsearch_fieldnames.test.ts.snap

@@ -79,6 +79,10 @@ export const METRIC_SYSTEM_FREE_MEMORY = 'system.memory.actual.free';
 export const METRIC_SYSTEM_TOTAL_MEMORY = 'system.memory.total';
 export const METRIC_SYSTEM_CPU_PERCENT = 'system.cpu.total.norm.pct';
 export const METRIC_PROCESS_CPU_PERCENT = 'system.process.cpu.total.norm.pct';
+export const METRIC_CGROUP_MEMORY_LIMIT_BYTES =
+  'system.process.cgroup.memory.mem.limit.bytes';
+export const METRIC_CGROUP_MEMORY_USAGE_BYTES =
+  'system.process.cgroup.memory.mem.usage.bytes';
 
 export const METRIC_JAVA_HEAP_MEMORY_MAX = 'jvm.memory.heap.max';
 export const METRIC_JAVA_HEAP_MEMORY_COMMITTED = 'jvm.memory.heap.committed';