New `storage-read-timeout` and `storage-write-timeout` cluster config options

[devanbenz]

ETA: 1.12.3

Two new configuration options have been added to the [cluster] section:

• storage-read-timeout (default: "15m") — Read inactivity timeout for Storage API RPC operations (StoreReadFilter, StoreReadGroup, StoreReadWindowAggregate). Resets on each read, so it detects stalled connections rather than slow operations.
• storage-write-timeout (default: "15m") — Write inactivity timeout for Storage API RPC operations. Same resettable behavior.

Both are applied to client and server sides of Storage API calls. When shard-reader-timeout or shard-writer-timeout is non-zero, the corresponding storage timeout is clamped to not exceed it.

Set to "0" to disable.

InfluxDB Enterprise 1.12.3+