Hi there, it is probably just some misunderstanding from my side but I am unable to visualise any data with Grafana from an InfluxDB database which is continuously filled via Telegraf. So far I can tell, the metrics are arriving in the database but I have the feeling that everything is simply sitting in some “cache” and therefore Grafana tells there is “no data”.
I am not sure how to debug this, but this is how one measurement looks like in the database:
> explain select * from cpu
QUERY PLAN
----------
EXPRESSION: <nil>
AUXILIARY FIELDS: cpu::tag, host::tag, usage_guest::float, usage_guest_nice::float, usage_idle::float, usage_iowait::float, usage_irq::float, usage_nice::float, usage_softirq::float, usage_steal::float, usage_system::float, usage_user::float
NUMBER OF SHARDS: 1
NUMBER OF SERIES: 17
CACHED VALUES: 43860
NUMBER OF FILES: 0
NUMBER OF BLOCKS: 0
SIZE OF BLOCKS: 0
The CACHED VALUES is increasing with every entry from Telegraf and I can see no errors in the logs, so that seems to work.
Any idea what to check next? Grafana is able to connect to the datasource and also recognises all the database query parameters correctly, but still, no data…
Is it maybe a misunderstanding and CACHED DATA is also “queryable”, or is that the source of error? If so, how can I decrease the cache size?
I have seen Reduce time for which influxdb holds data in cache - #3 by amoondra19 but there, the problem is the memory and not “missing data”.
