- Fix#743
- Rewrite InfluxDB storage for new InfluxDB API data structures.
- Store each measurement separately instead of storing all measurements
in a single big "table" with many columns/fields.
- Use tags add metadata to points, such as the container name. Tags are
a new feature in InfluxDB 0.9.
We should probably make memory as a cache type rather than storage.
RecentStats() can then be removed from the storage interface.
Will try it out as a separate PR.
is configurable via a flag and is set to 60 seconds by default.
The storage driver test code has been refactored a bit to help test the new buffering
feature and will be future proof.
Docker-DCO-1.1-Signed-off-by: Vishnu Kannan <vishnuk@google.com> (github: vishh)
We'll re-add new params as they are needed and as we understand influxdb
scalability better.
Docker-DCO-1.1-Signed-off-by: Rohit Jnagal <jnagal@google.com> (github: rjnagal)
For influxdb, we are planning to export lesser data. Having a separate
comparator will let us reuse the same tests.
Docker-DCO-1.1-Signed-off-by: Rohit Jnagal <jnagal@google.com> (github: rjnagal)
