[DocDB] Eliminate the need for explicit aggressive poll interval - wait_queue_poll_interval_ms for contentious workloads

[robertsami]

Jira Link: DB-5848

Description

We currently depend on a polling-based approach to resolve waiting transactions in the wait queue in order to achieve fairness under highly-contentious workloads, e.g. a workload where 10s of sessions are concurrently locking the same row.

Without aggressive polling (e.g. setting wait_queue_poll_interval_ms=5), such highly contentious workloads will suffer from high p99 latencies

Once we have https://github.com/yugabyte/yugabyte-db/issues/13578, we should ensure that highly contentious workloads can function with predictable p99 performance even with wait_queue_poll_interval_ms=100 or larger. Otherwise, we are trading off significant CPU overhead for fairness

Warning: Please confirm that this issue does not contain any sensitive information

• [X] I confirm this issue does not contain any sensitive information.

[robertsami]

for example, with the following transaction:

begin;
select * from foo where k = 1 for update;
commit;

if we have a single node local cluster on a 6 core machine, and we run this in 8 parallel threads, we will see something like 1/1000 requests experience 100x avg latency

[robertsami]

A tentative root cause has been identified, with a POC fix seemingly solving the issue. We have the following two code paths racing with each other:

Path 1 --

1. txn coordinator receives commit/abort for T1
2. txn coordinator sends UpdateTxn to participant for T1
3. participant calls SignalCommitted(T1)/SignalAborted(T1) on wait queue
4. wait queue signals all waiters currently known to be waiting on T1

Path 2 --

1. tserver receives WriteRequest for operation that is part of T2
2. write_query finds conflicts during conflict_resolution, including T1
3. conflict_resolution calls WaitQueue::WaitOn(waiter: T2, blockers: [T1])

step 3 of path 2 and step 4 of path 1 need to be synchronized, else T2 may miss the signal from the participant and be stuck in the wait queue until Poll() is called

[robertsami]

Depends on #21404