[Fleet] Fix some vars from preconfiguration not being added to package policies

[jen-huang]

Summary

Resolves elastic/fleet-server#742.

This PR fixes an issue where some variables defined in xpack.fleet.agentPolicies were not having their values correctly applied to the final package policies. This is seen on 7.15 Cloud deployments where the Fleet Server policies is missing values for max_connections and custom, even though they are specified in the preconfiguration setup.

This is due to a mistake in the ordering of objects that are merged in the helper function deepMergeVars. The override value should get merged last as, well, it is meant to override ;)

Testing

With Postman or cURL, send the below request to set up some preconfigured policy (I am using the exact Cloud setup in this case) prior to this PR. Observe the policy in Fleet UI, notice the empty fields:

After applying this PR, run the request again after incrementing the id. The new policy should have the fields correctly filled out:

PUT /api/fleet/setup/preconfiguration
{
  "agentPolicies": [
    {
      "id": "some-test-1",
      "name": "Some test 1",
      "description": "Default agent policy for agents hosted on Elastic Cloud",
      "is_default": false,
      "is_managed": true,
      "is_default_fleet_server": false,
      "namespace": "default",
      "monitoring_enabled": [],
      "unenroll_timeout": 600,
      "package_policies": [
        {
          "name": "Fleet Server",
          "package": {
            "name": "fleet_server"
          },
          "inputs": [
            {
              "type": "fleet-server",
              "keep_enabled": true,
              "vars": [
                {
                  "name": "host",
                  "value": "0.0.0.0",
                  "frozen": true
                },
                {
                  "name": "port",
                  "value": 8220,
                  "frozen": true
                },
                {
                  "name": "max_connections",
                  "value": 200
                },
                {
                  "name": "custom",
                  "value": "cache:\n  num_counters: 2000        # Limit the size of the hash table to rougly 10x expected number of elements\n  max_cost: 2097152         # Limit the total size of data allowed in the cache, 2 MiB in bytes.\nserver.limits:\n   policy_throttle: 200ms  # Roll out a new policy every 200ms; roughly 5 per second.\n   checkin_limit:\n     interval: 50ms        # Check in no faster than 20 per second.\n     burst: 25             # Allow burst up to 25, then fall back to interval rate.\n     max: 100              # No more than 100 long polls allowed. THIS EFFECTIVELY LIMITS MAX ENDPOINTS.\n   artifact_limit:\n     interval: 100ms       # Roll out 10 artifacts per second\n     burst: 10             # Small burst prevents outbound buffer explosion.\n     max: 10               # Only 10 transactions at a time max.  This should generally not be a relavent limitation as the transactions are cached.\n   ack_limit:\n     interval: 10ms        # Allow ACK only 100 per second.  ACK payload is unbounded in RAM so need to limit.\n     burst: 20             # Allow burst up to 20, then fall back to interrval rate.\n     max: 20               # Cannot have too many processing at once due to unbounded payload size.\n   enroll_limit:\n     interval: 100ms       # Enroll is both CPU and RAM intensive.  Limit to 10 per second.\n     burst: 5              # Allow intial burst, but limit to max.\n     max: 10               # Max limit.\nserver.runtime:\n  gc_percent: 20          # Force the GC to execute more frequently: see https://golang.org/pkg/runtime/debug/#SetGCPercent\n"
                }
              ]
            }
          ]
        }
      ]
    }
  ]
}

[elasticmachine]

Pinging @elastic/fleet (Team:Fleet)

[kibanamachine]

💚 Build Succeeded

• continuous-integration/kibana-ci/pull-request
• Commit: 4cbd5f3
• Storybooks Preview

Metrics [docs]

✅ unchanged

To update your PR or re-run it, just comment with:
@elasticmachine merge upstream

cc @jen-huang

[juliaElastic]

LGTM

[joshdover]

Change LGTM. Can we open an issue for adding test coverage for this?

[kpollich]

Change LGTM. Can we open an issue for adding test coverage for this?

Created #113248 to capture test coverage. I will take this on.

[kibanamachine]

💚 Backport successful

Status	Branch	Result
✅	7.x
✅	7.15

The backport PRs will be merged automatically after passing CI.

[dikshachauhan-qasource]

Hi @kpollich

We have validated this PR And found this PR is fixed now in 8.0 snapshot, 7.15.1 snapshot and 7.16 snapshot build.

• Observation now default settings are available under Fleet server Integration on Cloud deployment.

Screenshots:

Hence, it is working fine now.

Thanks
QAS