Track epic issues count on usage data

What does this MR do?

Track epic issues with usage ping.

related to #233156 (closed)

Queries with plans

Gitlab::UsageData.count(EpicIssue)
   (2.6ms)  SELECT MIN("epic_issues"."id") FROM "epic_issues"
   (0.6ms)  SELECT MAX("epic_issues"."id") FROM "epic_issues"
   (0.3ms)  SELECT COUNT("epic_issues"."id") FROM "epic_issues" WHERE "epic_issues"."id" >= $1 AND "epic_issues"."id" < $2  [["id", 2], ["id", 342996]]


EXPLAIN ANALYZE SELECT MIN("epic_issues"."id") FROM "epic_issues" \g
                                                                       QUERY PLAN                        
                                                
---------------------------------------------------------------------------------------------------------
------------------------------------------------
 Result  (cost=0.45..0.46 rows=1 width=4) (actual time=1.806..1.806 rows=1 loops=1)
   InitPlan 1 (returns $0)
     ->  Limit  (cost=0.42..0.45 rows=1 width=4) (actual time=1.803..1.803 rows=1 loops=1)
           ->  Index Only Scan using epic_issues_pkey on epic_issues  (cost=0.42..6780.40 rows=254733 wid
th=4) (actual time=1.802..1.802 rows=1 loops=1)
                 Index Cond: (id IS NOT NULL)
                 Heap Fetches: 0
 Planning Time: 1.390 ms
 Execution Time: 1.980 ms
(8 rows)



EXPLAIN ANALYZE SELECT MAX("epic_issues"."id") FROM "epic_issues" \g
                                                                            QUERY PLAN                   
                                                         
---------------------------------------------------------------------------------------------------------
---------------------------------------------------------
 Result  (cost=0.45..0.46 rows=1 width=4) (actual time=0.887..0.887 rows=1 loops=1)
   InitPlan 1 (returns $0)
     ->  Limit  (cost=0.42..0.45 rows=1 width=4) (actual time=0.882..0.882 rows=1 loops=1)
           ->  Index Only Scan Backward using epic_issues_pkey on epic_issues  (cost=0.42..6780.40 rows=2
54733 width=4) (actual time=0.881..0.881 rows=1 loops=1)
                 Index Cond: (id IS NOT NULL)
                 Heap Fetches: 1
 Planning Time: 0.105 ms
 Execution Time: 0.907 ms
(8 rows)


EXPLAIN ANALYZE SELECT COUNT("epic_issues"."id") FROM "epic_issues" WHERE "epic_issues"."id" >= 2 AND "epic_issues"."id" < 342996

---------------------------------------------------------------------------------------------------------
------------------------------------------------------------------
 Finalize Aggregate  (cost=7196.82..7196.83 rows=1 width=8) (actual time=96.077..96.077 rows=1 loops=1)
   ->  Gather  (cost=7196.61..7196.82 rows=2 width=8) (actual time=91.537..106.051 rows=3 loops=1)
         Workers Planned: 2
         Workers Launched: 2
         ->  Partial Aggregate  (cost=6196.61..6196.62 rows=1 width=8) (actual time=88.347..88.348 rows=1
 loops=3)
               ->  Parallel Index Only Scan using epic_issues_pkey on epic_issues  (cost=0.42..5931.26 ro
ws=106138 width=4) (actual time=0.749..82.218 rows=85048 loops=3)
                     Index Cond: ((id >= 2) AND (id < 342996))
                     Heap Fetches: 50870
 Planning Time: 0.220 ms
 Execution Time: 106.171 ms
(10 rows)

Detailed explain

https://explain.depesz.com/s/Oesv https://explain.depesz.com/s/QkTR
https://explain.depesz.com/s/44Rb