Bug #2323
Grafana has high CPU usage and high I/O
| Status: | Resolved | Start date: | 06/14/2015 | |
|---|---|---|---|---|
| Priority: | Normal | Due date: | ||
| Assignee: | J. Wienke | % Done: | 100% | |
| Category: | integration | |||
| Target version: | 2.0 |
Description
InfluxDB uses ~600% CPU for about 1 - 1/2 hours. Also there is a high I/O during this time (4-6MB).
Is this intended? I suppose it is self-optimisation of the database (e.g. index generation)... If so, can this be forced to happen only at night (eg. past 1 am)?
see here for an image of the usage (CSRA internal)
History
#1 Updated by J. Wienke over 9 years ago
- Category set to integration
#2 Updated by J. Wienke over 9 years ago
- Status changed from New to In Progress
- % Done changed from 0 to 40
Basically we are limited by the integration between influxdb and grafana and the (in-) ability to use continuous downsampling queries. I have informally evaluated graphite as a different backend again and performance is much better here. Usable support is implemented in the master tree.
#3 Updated by J. Wienke about 9 years ago
- Status changed from In Progress to Resolved
- Assignee set to J. Wienke
- Target version set to 2.0
- % Done changed from 40 to 100
This is basically solved with switching to graphite.