query too slow

version hasura/graphql-engine:v1.3.3.cli-migrations-v2

```
{
"operationName": "user_activity_logs",
"query": "query user_activity_logs($distinct_on: [user_activity_logs_select_column!], $limit: Int, $offset: Int, $order_by: [user_activity_logs_order_by!], $where: user_activity_logs_bool_exp) {\n items: user_activity_logs(limit: $limit, distinct_on: $distinct_on, offset: $offset, order_by: $order_by, where: $where) {\n id\n user_id\n action\n created_at\n project_id\n specific_name\n details\n module\n sub_module\n type\n updated_at\n user {\n name\n email\n __typename\n }\n moduleByModuleId {\n module\n __typename\n }\n moduleBySubModuleId {\n module\n __typename\n }\n __typename\n }\n items_aggregate: user_activity_logs_aggregate(where: $where) {\n aggregate {\n count\n __typename\n }\n __typename\n }\n}\n",
"variables": {
"limit": 5,
"offset": 0,
"order_by": [
{
"created_at": "desc"
}
],
"where": {
"_and": [
{},
{},
{},
{},
{}
]
}
}
}
```
that query took 53 sec to execute, if i execute `select * from user_activity_logs order by created_at desc limit 5` directly in the database takes no more than 1sec, how can i track the problem?

Start by reproducing the supplied user_activity_logs GraphQL operation and comparing its execution with the direct PostgreSQL query shown in the report. Trace the request and its generated database work to explain the 53-second delay; done means the cause is identified and a targeted fix or actionable diagnosis is recorded.

Written by the indexing model from the issue text.