v1.8.10 out of memory

A long-running Influxd ran out of memory on a 4GB RAM VPS, apparently after executing a slow query with `LIMIT 1`. The query had execute fine 15 times since influxd was started ~4 days earlier with `influxd -config /etc/influxdb/influxdb.conf 2>&1 | tee influxd.log`.

__Environment info:__

* System info: Linux 4.15.0-166-generic x86_64
* InfluxDB version: v1.8.10 (git: 1.8 688e697c51fd)
* Other relevant environment details: no container; 100G xfs disk, 9G free

__Config:__
[influxdb.conf.txt](https://github.com/influxdata/influxdb/files/8105141/influxdb.conf.txt)

__Logs:__
[error log](https://github.com/influxdata/influxdb/files/8105130/OOM_2022-02-18.txt)

__Performance:__
Unfortunately, as I wasn't anticipating the OOM, I didn't run the profiling commands while the bug was actively happening.

Start by reviewing the linked OOM error log and influxdb.conf.txt, then investigate the slow query run through the influxd command described in the report. The issue is complete when the memory exhaustion can be reproduced or explained and a verified resolution prevents this long-running query from exhausting the available RAM.

Written by the indexing model from the issue text.