Neither DataFrameClient nor InfluxDBClient will return more than 10k results

[ghost]

I've tried altering the number of results InfluxDB itself returns in the query string; this didn't work either. I notice theres a chunked param to the query methods, but these are not documented and a quick look at the source leads me to believe they're unused?

[pkittenis]

Have implemented chunked responses in this PR. The other option is setting max-row-limit to 0 in influxdb config - this may cause DB to run out of memory.

Not sure if this project is actively maintained though, will have to wait and see. In the meantime you can install the PR branch directly if you'd like.

Have need of it also for an influxdb graphite client, may have to point requirements to the fork instead of the official client.

[ghost]

I'll need portability so it's best that I stick with what's in pip. Thanks for the status update, though. If you know of any alternative (specifically, getting huge amounts of data from influxdb into a data frame easily), I'd love to hear about it.

Thanks again

[pkittenis]

For what it's worth, pip and its requirements.txt can refer to git repos directly, including branches, PRs, tags and so on.

Not aware of any other options.

[ghost]

Thanks for the info

[saroele]

I have the same issue, but only since we upgraded influxdb from 1.0 to 1.2. I'm using influxdb-python 4.0.0.

[saroele]

I ammended the PR of pkittenis so the DataFrameClient now also correctly uses chunking.
Tested, works for me.