These are chat archives for influxdata/influxdb

23rd
Mar 2017
Sam
@shamsalmon
Mar 23 2017 00:26
hello
Ivan Vaskevych
@IvanVas
Mar 23 2017 16:23
Hi guys! I need to store sensor measurements, there are about 10 types and 1k different sensor. I'll be writing them separately (sensor, type, value) and reading mostly together (sensor, * types)
What's the best structure to do that? Separate "measurements" per my type, or everything in one? Or, maybe, "measurement" per sensor ID?
Jack Zampolin
@jackzampolin
Mar 23 2017 18:22
@vladimirshin Yes it is! A single server can handle upwards of 500k metrics/sec
@manojgour Please post your question over in community.influxdata.com
@nnungest I'm working on a sandbox on top of docker compose https://github.com/jackzampolin/sandbox is you are interested.
@csaurav Thats the best documentation out there
@mani214 Up your open file handles limit ulimit -n 65000. This might mean you have too many databases/shards. If you could post any other questions over at community.influxdata.com
Jack Zampolin
@jackzampolin
Mar 23 2017 18:29
@irfn You need to store 10M field values https://docs.influxdata.com/influxdb/v1.2/concepts/glossary/#field across 1M series https://docs.influxdata.com/influxdb/v1.2/concepts/glossary/#series? A single server of influx can handle between 5-10M series and 500k+ writes/sec.
@r3cgm :wave:
Jack Zampolin
@jackzampolin
Mar 23 2017 18:36
@shamsalmon :wave: