panic in bolt/metrics.go due to concurrent map iteration and map write

[pradig]

Every couple of days my influxdb dies with a panic (e.g. after 13452 or 236 or 6725) minutes.

Linux orfeo 5.4.0-65-generic #73-Ubuntu SMP Mon Jan 18 17:25:17 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux

ts=2021-03-18T09:19:17.915373Z lvl=info msg="Welcome to InfluxDB" log_id=0Sy9xmbl000 version=2.0.4 commit=4e7a59bb9a build_date=2021-02-08T17:47:02Z

Prometheus is at version 2.25.0

fatal error: concurrent map iteration and map write
goroutine 20561008 [running]:
runtime.throw(0x279ee19, 0x26)
/usr/local/go/src/runtime/panic.go:1116 +0x72 fp=0xc049e89bf0 sp=0xc049e89bc0 pc=0x438af2
runtime.mapiternext(0xc049e89e68)
/usr/local/go/src/runtime/map.go:853 +0x554 fp=0xc049e89c70 sp=0xc049e89bf0 pc=0x412134
runtime.mapiterinit(0x237bde0, 0xc0635042a0, 0xc049e89e68)
/usr/local/go/src/runtime/map.go:843 +0x1c5 fp=0xc049e89c90 sp=0xc049e89c70 pc=0x411ae5
github.com/influxdata/influxdb/v2/bolt.(*Client).Collect(0xc00186e000, 0xc065192240)
/home/circleci/go/src/github.com/influxdata/influxdb/bolt/metrics.go:223 +0x835 fp=0xc049e89ed8 sp=0xc
049e89c90 pc=0x1008995
github.com/prometheus/client_golang/prometheus.(*Registry).Gather.func1()
/home/circleci/go/pkg/mod/github.com/prometheus/client_golang@v1.5.1/prometheus/registry.go:443 +0x1a2
fp=0xc049e89fe0 sp=0xc049e89ed8 pc=0xb5b2a2
runtime.goexit()
/usr/local/go/src/runtime/asm_amd64.s:1374 +0x1 fp=0xc049e89fe8 sp=0xc049e89fe0 pc=0x471821
created by github.com/prometheus/client_golang/prometheus.(*Registry).Gather
/home/circleci/go/pkg/mod/github.com/prometheus/client_golang@v1.5.1/prometheus/registry.go:454 +0x5ce