TODO

[aviau]

TODO LIST

• Review the docs
• Update and test all the tutorials
• Fix the DataFrameClient and improve coverage (see DataFrameClient with 0.9.0 #108 )
• Implement more operations (user management, etc.)

DONE

• Debug UDP support (I couldn't get it working)
• Fix the SeriesHelper
• Support writing in batch (see batch_size logic in write_points too soon leaves the loop #102)

[dashesy]

I have a small working library that plan to make it into Pandas, you can take a look if you want here

[aviau]

@dashesy Would you keep me updated on that please? If this is ever merged in pandas I would love to support it in the lib.

[areski]

The test coverage on DataFrameClient is quite weak, it might be nice to add this to the TODO.

[aviau]

Noted.

[cannium]

As I noticed, InfluxDB v0.9 supports clustering. Does influxdb-python supports connecting to multiple hosts(a cluster) now? I think the new version should handle clustering properly.

[gst]

about :

Debug UDP support (I couldn't get it working)

it's (very "simply", with write_points() method) tested in last version.. within class UdpTests.. which could probably be extended with others udp tests more advanced, eventually.

[cannium]

I posted a pull request for clustering. Please review.

[gst]

@ReAzem
I was still asking myself why you tagged "so quickly" and more important : why you taggeg with v1.0.0.. (and then after 1.0.1 & 1.0.2)..

we could have had/followed the same (at least major) version than the server.. now it's over unless we delete the tags.. (which I'd vote for but it depends..)

wdyt ?

[aviau]

@gst

The versioning of the lib has nothing to do with the versioning of the server. This is not the same project.

I am following the Semantic Versioning spec.

major.minor.patch

Everytime the API breaks: major release
Every new feature: minor release
Every bug patch/fix: patch release

The next tag will be 2.0.0 because we broke the API with ResultSet.

[gst]

The versioning of the lib has nothing to do with the versioning of the server. This is not the same project.

that is a shame ! ;) no kidding ;) Ofcourse there aren't the same project :p but they are quite tightly coupled.. aren't they ?
I find that it's quite often that a client follow the same version (at least major) than the server for which it is the client .. just for the sake of clarity / simplicity..
examples : mysql , postgres , others..

then why did you raised from v0.4.1 to v1.0.0 .. ? we certainly could have had v0.9-rcXX :s

The next tag will be 2.0.0 because we broke the API with ResultSet.

because you already tagged v1.0.0 previously ;) which I still don't see a/the reason actually..

[aviau]

I find that it's quite often that a client follow the same version (at least major) than the server for which it is the client

Can't do that and respect semver at the same time. I think semver is more important than this.

then why did you raised from v0.4.1 to v1.0.0 .. ? we certainly could have had v0.9-rcXX :s

I wanted to release to PyPI ASAP because of the influxdb08 sub-module. This was a good time so I released 1.0.0. 1.0.0 should have been released a long time ago, I didn't manage versioning at this time.

See the Semver Specification:

How do I know when to release 1.0.0?

If your software is being used in production, it should probably already be 1.0.0. If you have a stable API on which users have come to depend, you should be 1.0.0. If you're worrying a lot about backwards compatibility, you should probably already be 1.0.0.

[gst]

ok for the Semver spec..

but hell :

How do I know when to release 1.0.0?
If your software is being used in production, it should probably already be 1.0.0.

:| there are people using influxdb-server (but v0.8 then...) and its python-client in production (which should then be influxdb-python v0.4.1 ....) ??

[aviau]

:| there are people using influxdb-server (but v0.8 then...) and its python-client in production (which should then be influxdb-python v0.4.1 ....) ??

Yes. See here (hopefully it all worked out for him!):

/pull/99
ChristopherRabotin:
I hope to start using influxdb in prod starting the middle of next week. :D

InfluxDB 0.5.0 was announced as ready for production: http://influxdb.com/blog/2014/03/24/influxdb_v0_5_0_ready_for_production.html

I will absolutely keep respecting the semver spec. I will also never revert a tag. PyPI won't let me do it and its just terrible practice anyways.

[ChristopherRabotin]

I confirm, if need be, that I use influxdb python in production.

On Mon, Apr 13, 2015, 21:01 Alexandre Viau [email protected] wrote:

:| there are people using influxdb-server (but v0.8 then...) and its
python-client in production (which should then be influxdb-python v0.4.1
....) ??

Yes:

#99
#99
ChristopherRabotin:
I hope to start using influxdb in prod starting the middle of next week. :D

I will absolutely keep respecting the semver spec. I will also never
revert a tag. PyPI won't let me do it and its just terrible practice
anyways.

—
Reply to this email directly or view it on GitHub
#109 (comment)
.

[gst]

points taken.

PyPI won't let me do it and its just terrible practice anyways

yeah I know.. PyPI is just right in that.

[peterchou91]

i got a problem ,when i ran an example:
1 from influxdb.influxdb08 import InfluxDBClient
2
3 json_body = [{"name": "cpu_load_short","timestamp": "2009-11-10T23:00:00Z","name":"1"}]
4 client = InfluxDBClient('10.136.108.128', 8086, 'root', 'root', 'example')
5 client.create_database('example')
6 client.write_points(json_body)
7 result = client.query('select * from cpu_load_short;')
8 print("Result: {0}".format(result))
when the program above finished,it prints:
Traceback (most recent call last):
File "test.py", line 8, in
result = client.query('select * from cpu_load_short;')
File "/usr/lib/python2.7/site-packages/influxdb/influxdb08/client.py", line 419, in query
chunked=chunked)
File "/usr/lib/python2.7/site-packages/influxdb/influxdb08/client.py", line 447, in _query
expected_response_code=200
File "/usr/lib/python2.7/site-packages/influxdb/influxdb08/client.py", line 255, in request
raise InfluxDBClientError(response.content, response.status_code)
influxdb.influxdb08.client.InfluxDBClientError: 400: Couldn't find series: cpu_load_short

this is so wired,i can't figure it out,can anybody help me ?
thanks

[ChristopherRabotin]

@peterchou91 you're attempting to read from a time series which does not exist. If you think it exists, can you check that it is indeed called cpu_load_short? To create that time series, all you need to do is write points to it.

BTW, I'm not sure here is the appropriate place to discuss your issue.

[cp-richard]

Any reason not to be using this module with 0.9.x now ?

[aviau]

@richard-kti The module supports 9.0.x.

[brewneaux]

Should this be updated to use line instead of JSON?

[aviau]

Should this be updated to use line instead of JSON?

Already done as of 2.7.0

[brewneaux]

I must have missed that, my apologies.

[aviau]

I was using this issue when doing the transition to 0.9.0 and taking over the repository. I don't think it is useful anymore, as it is turning into a forum. I'll close this.