Periodically very high memory and CPU time usage


I am using InfluxDB 1.7.9 to store some data on a windows system. I also use Telegraf to monitor the working set of the service. I recognised, that every 24 to 26 hours the memory usage as well as the processor time of the InfluxDB service increases dramatically.

I was just wondering if there is anything I can do about this.

Any help would be highly appreciated!


I’m assuming your shard duration is roughly a day. These spikes are shard compactions and are required for healthy operation of your InfluxDB instance

1 Like


thank you for your response.
Can I reduce the height of these spikes by reducing the shard duration?

can you solve the problem?