InfluxDB only returns partial data. I suspect that InfluxDB has an internal limit on the data returned. Is there any documentation explaining this behavior?
The total number of points is 31,622,400, but InfluxDB only returned 937,462 points.
The query using github.com/influxdata/influxdb-client-go/v2 also behaves similarly, with the number of points returned (approximately 7,900,000) varying each time.
I am using InfluxDB v2.7.10 with the image influxdb:2-alpine. The host system information is:
$ uname -r -v
6.5.0-41-generic #41~22.04.2-Ubuntu SMP PREEMPT_DYNAMIC Mon Jun 3 11:32:55 UTC 2
I had a very similar issue. Try looking in your telegraf logs for: Metric buffer overflow. If that is the problem you can increase it, however, 31 million is a ton. You may need to break it apart into separate polls instead of just moving your buffer limit to 32 million.
Thanks for the response. I’m not using Telegraf. It seems the issue might be due to the data volume being too large, so I’ll try querying it in smaller parts.
After setting the HTTP request timeout for influxdb-client-go, all the data are returned. The default request timeout for influxdb-client-go is 20 seconds.