Add kubernetes_agents usage metric

What does this MR do?

Related issue: #219716 (closed)

SQL:

[1] pry(main)> Gitlab::UsageData.count(::Clusters::Agent)
   (4.1ms)  SELECT MIN("cluster_agents"."id") FROM "cluster_agents"
   (0.3ms)  SELECT MAX("cluster_agents"."id") FROM "cluster_agents"
   (0.2ms)  SELECT COUNT("cluster_agents"."id") FROM "cluster_agents" WHERE "cluster_agents"."id" BETWEEN $1 AND $2  [["id", 1], ["id", 100000]]
=> 1

#database-lab plan:

explain SELECT COUNT("cluster_agents"."id") FROM "cluster_agents" WHERE "cluster_agents"."id" BETWEEN 1 AND 100000;

 Aggregate  (cost=7.74..7.75 rows=1 width=8) (actual time=0.044..0.044 rows=1 loops=1)
   Buffers: shared hit=4
   ->  Index Only Scan using cluster_agents_pkey on public.cluster_agents  (cost=0.15..7.73 rows=4 width=8) (actual time=0.040..0.040 rows=0 loops=1)
         Index Cond: ((cluster_agents.id >= 1) AND (cluster_agents.id <= 100000))
         Heap Fetches: 0
         Buffers: shared hit=4

Time: 0.290 ms
  - planning: 0.131 ms
  - execution: 0.159 ms
    - I/O read: 0.000 ms
    - I/O write: 0.000 ms

Shared buffers:
  - hits: 4 (~32.00 KiB) from the buffer pool
  - reads: 0 from the OS file cache, including disk I/O
  - dirtied: 0
  - writes: 0

Does this MR meet the acceptance criteria?

Conformity

• Changelog entry
• [-] Documentation (if required)
• Code review guidelines
• Merge request performance guidelines
• Style guides
• Database guides
• [-] Separation of EE specific content

Availability and Testing

• Review and add/update tests for this feature/bug. Consider all test levels. See the Test Planning Process.