unexpected fault address resulted in 7 day and 2 hour data loss

__Steps to reproduce:__
¯\\\_(ツ)\_/¯

1. run influxdb for a few months and hope it happens to you?

note: i've had this happen before
i think it was always a loss of ~7 days, and always not the most recent 7 days
(just looking back, i found a loss from 2022-03-06 23:00:00 to 2022-03-14 02:00:00 on another bucket)
i think i had more than these two, maybe something got recovered automatically?

__Expected behavior:__
influxdb to either not crash, or crash but not lose data

__Actual behavior:__
influxdb, apparently spontaneously, decided to segfault
it lost 7 days and 2 hours of data in some measurements in 1 bucket
it lost this for a period approximately 4 days before the crash
(so going backwards in time, i have:
today's data up to the crash (2022-12-30 02:46:50)
data ~4 days before the crash (2022-12-26 01:00:00)
7 days and 2 hours missing (2022-12-18 23:36:20)
older data is still there)

__Environment info:__

* System info: Linux 6.1.0 x86_64
* NixOS unstable
* InfluxDB version: InfluxDB 2.5.1 (git: v2.5.1) build_date: 2022-12-30T20:33:07Z
* reported build_date is time of executing `influxd version`

__Config:__
apparently nothing non-stock, except a mistyped attempt to disable telemetry reporting

__Logs:__
[influxdb2-crash-trimmed.log](https://github.com/influxdata/influxdb/files/10326174/influxdb2-crash-trimmed.log)

Begin with the attached influxdb2-crash-trimmed.log and the reported InfluxDB 2.5.1 environment on Linux 6.1.0 x86_64 and NixOS unstable. Determine what caused the segfault and the missing-data interval; done means the crash and data loss are explained and prevented.

Written by the indexing model from the issue text.