[Bug] InfluxDB storage backend falls behind and loses data

[robagar]

Describe the bug

Under moderately heavy load, the InfluxDB v1 backend falls behind and eventually starts dropping samples.

See https://discordapp.com/channels/914168414178779197/1207326913186762794 for the details

To reproduce

1. Configure router to store in InfluxDB
2. Start publisher for any data at a high-ish frequency, say 100Hz
3. Start publisher at 1Hz for events with a counter
4. Wait...
5. Make a get request using a time range for the events

Once the storage falls behind by ~1m 40s, it starts dropping samples in batches, not making PUT requests to InfluxDB for them

System info

zenohd v0.10.1-rc-1-g15b36a0f built with rustc 1.72.0 (5680fa18f 2023-08-23)
InfluxDB 1.8.10

[robagar]

The InfluxDB docs recommend sending data in batches for performance, so here's our solution

https://github.com/engineeredarts/zenoh-backend-influxdb/tree/TR-6326-zenoh-influxdb

• Collect PUT queries and send in single request if put_batch_size is set in the storage volume config
• Send when batch is full, or put_batch_timeout_ms milliseconds after the first one, default 1000ms

Let me know if you want a PR :)