I'm working on a monitoring system to retrieve hardware metric by telegraf and store the time-series data to TDengine.
But seem the data insertion is very slow even making telegraf reject to work anymore.
I got error msg like following:
2021-12-28T11:36:14Z W! [agent] ["outputs.http"] did not complete within its flush interval
2021-12-28T11:36:23Z E! [agent] Error writing to outputs.http: Post "http://xxxxxx:6041/influxdb/v1/write?db=bmc": context deadline exceeded (Client.Timeout exceeded while awaiting headers)
2021-12-28T11:36:35Z W! [agent] ["outputs.http"] did not complete within its flush interval
2021-12-28T11:36:43Z E! [agent] Error writing to outputs.http: Post "http://xxxxxx:6041/influxdb/v1/write?db=bmc": context deadline exceeded (Client.Timeout exceeded while awaiting headers)
2021-12-28T11:37:02Z W! [agent] ["outputs.http"] did not complete within its flush interval
2021-12-28T11:37:07Z E! [agent] Error writing to outputs.http: Post "http://xxxxxx:6041/influxdb/v1/write?db=bmc": context deadline exceeded (Client.Timeout exceeded while awaiting headers)
2021-12-28T11:37:08Z W! [outputs.http] Metric buffer overflow; 7824 metrics have been dropped
2021-12-28T11:37:21Z W! [agent] ["outputs.http"] did not complete within its flush interval
2021-12-28T11:37:28Z E! [agent] Error writing to outputs.http: Post "http://xxxxxxxx:6041/influxdb/v1/write?db=bmc": context deadline exceeded (Client.Timeout exceeded while awaiting headers)
2021-12-28T11:37:28Z W! [outputs.http] Metric buffer overflow; 14363 metrics have been dropped
Does anyone know how to tune TDengine to work better performance?
Edit /etc/taos/taosadatper.toml and add following lines:
debug = true
port = 6041
logLevel = "debug"