We don't need these as we have Sidekiq-specific metrics. Having these
metrics with Sidekiq labels also increases cardinality.

We keep the DB counters in logs for Sidekiq, but remove them from
metrics too.

The plan for RSpec 4.0 is to disable monkey patching, reference:
https://rspec.info/blog/2013/07/the-plan-for-rspec-3/#zero-monkey-patching-mode
https://relishapp.com/rspec/rspec-core/docs/configuration/zero-monkey-patching-mode

This commit stops using RSpec monkey patching in several specs.

It introduces detailed logs for each Redis instance.

Before this change, we used to have just the following
Redis logs for API, Sidekiq and Web requests:

- redis_calls
- redis_duration_s
- redis_read_bytes
- redis_write_bytes

Here we introduce the following (when > 0):

- redis_action_cable_calls
- redis_action_cable_duration_s
- redis_action_cable_read_bytes
- redis_action_cable_write_bytes
- redis_cache_calls
- redis_cache_duration_s
- redis_cache_read_bytes
- redis_cache_write_bytes
- redis_queues_calls
- redis_queues_duration_s
- redis_queues_read_bytes
- redis_queues_write_bytes
- redis_shared_state_calls
- redis_shared_state_duration_s
- redis_shared_state_read_bytes
- redis_shared_state_write_bytes

That's valuable information about how GitLab
use Redis instances.

This adds instrumentation for the number of bytes written to and read
from Redis. To avoid having to monkey-patch multiple methods of
Redis::Client, this change introduces a custom Redis driver which
inherits from the default Redis::Connection::Ruby driver.

Counts include the size of the commands (GET, SET etc.), but exclude
the size overhead of Redis serialization (type, length prefix,
trailing \r\n).

https://gitlab.com/gitlab-com/gl-infra/scalability/-/issues/318

We have 6 (microsecond) precision for a few Go service
timings, so making all existing *_duration_s on
Rails/API/Sidekiq use a 6 decimal precision instead of 2
would make more sense, and that's what we accomplish here.

It makes the decision on how to log timings within
JSON logs based on
https://www.robustperception.io/who-wants-seconds.

It has been difficult to see trends in the number of Redis calls without
having the number and duration of the calls in structured logs. This
commit adds `redis_calls` and `redis_duration_ms` fields to all relevant
logs (e.g. api_json.log, production_json.log, Sidekiq, etc.).

Closes https://gitlab.com/gitlab-org/gitlab/issues/208821

Adds a Prometheus histogram, `sidekiq_jobs_queue_duration_seconds` for
recording the duration that a Sidekiq job is queued for before being
executed.

This matches the scheduling_latency_s field emitted from structured
logging for the same purpose.