I have several instances of telegraf 1.1.2 running on a FreeBSD 10.3 servers (real hardware). After several days, they seem to be dropping metrics.
For example, the following chart on grafana shows many gaps for events that are always happening in my system. These are collected using the statsd component. I also have similar gaps for internal data collection (CPU, memory load, etc.) <https://lh3.googleusercontent.com/-44DR9wO6GMo/WG0RK2qx4HI/AAAAAAABp-k/SMcSi3kG7TQzHJdJGNtIJaHPNdiVezSCgCLcB/s1600/Screen%2BShot%2B2017-01-04%2Bat%2B10.13.25.png> At 9:44 today I restarted the telegraf process and there were no more gaps in the data. This shows me that it is indeed telegraf, not influxdb, grafana, or the statsd clients. None of the machines were logging anything into the telegraf log. There were a couple of "unable to store data" warnings from a week ago, though, but there were only one or two of those spread out over a period of days. There were also occasional "data collection took more than 10s for 10s interval" but again, those were several days ago as well. Before I restarted the processes, I checked if they were consuming lots of memory and they were not. After restart they were smaller, though by about 20MB. The servers overall are not starved for memory or CPU resources. What else might I look for here to see why telegraf becomes unreliable and inconsistent after several days? Some of the telegrafs were running since Dec. 30, and others since December 17. -- Remember to include the version number! --- You received this message because you are subscribed to the Google Groups "InfluxData" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/influxdb. To view this discussion on the web visit https://groups.google.com/d/msgid/influxdb/87b5e70e-b762-4f22-9857-7e94694c90d6%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
