For the full commit log, see here.
- Ensure workers can send data popped off the queue at shutdown (#239, @onlynone)
- Add support for special characters in password when using url keyword (#242, @swistak35)
- Add Ruby 3.0 support (#249, @dentarg, @paul and @track8)
- Support characters that need quoting for usernames and database names (#248, @estheruary)
- Allow dropping of specific series from specific DBs (#233, @cantino)
- Add support for MRI 2.7 (#235, @jeffkowalski)
- Raise a LineProtocolError if attempting to write empty values as field set is required. This adds descriptive feedback to the error "unable to parse '{series},{tags} ': invalid field format" (#236, @todtb)
- Add support for configuring HTTP Proxy (#238, @epchris)
- Drop support for Ruby 2.2, since Bundler dropped it and we want to use Bundler in the development cycle as well.
- Fix issue with tag values ending in a backslash.
- Fix newly introduced
InfluxDB.now(precision)
for precisions larger than "s".
- Added
InfluxDB.now(precision)
andInfluxDB::Client#now
as companions toInfluxDB.convert_timestamp
.
- Added
InfluxDB.convert_timestamp
utility to convert aTime
to a timestamp with (taken from PR influxdb-rails#53 by @ChrisBr).
- Fix
InfluxDB::Client#delete_retention_policy
: the database name argument is now quoted (#221, #222 @AishwaryaRK). - Add
InfluxDB::Client#list_measurements
and#delete_measuerment
(#220)
-
Add batch query support via
InfluxDB::Client#batch
(and usingInfluxDB::Query::Batch
). Using multiple queries joined with;
will cause issues withClient#query
in combination with eitherGROUP BY
clauses or empty results, as discussed in #217.Initial code and PR#218 from @satyanash.
- Fix
NoMethodError
inInfluxDB::Client#list_retention_policies
when the database has no RPs defined (#213, @djoos)
- Add async option to block on full queue (#209, @davemt)
- Add support for
SHARD DURATION
in retention policy (#203, @ljagiello)
-
Add support for precision, retention policy and database parameters to async writer (#140, #202 @rockclimber73)
Attention You may need to validate that your calls to the write API (
InfluxDB::Client#write
,#write_point
,#write_points
) don't accidentally included a precision, RP, and/or database argument. These arguments were ignored until now. This is likely if you have changed your client instance configuration in the past, when you added aasync: true
. Updating might cause data inconsistencies!
- Bugfix in
InfluxDB::PointValue
: Properly encode backslashes (#200)
- Bugfix in async client: Flush queue before exit (#198, #199 @onlynone)
- Dropped support for Ruby < 2.2.
- Updated dependencies.
- Refactor some method declarations, to take kwargs instead of an options hash (this shouldn't break call sites).
- Allow configuration by an URL (idea by @carlhoerberg in #188).
- Improved logging (#180).
- (Backport from v0.4.1) Bugfix in async client: Flush queue before exit (#198, #199 @onlynone)
- (Backport from v0.4.2) Bugfix in
InfluxDB::PointValue
: Properly encode backslashes (#200)
- This is propably the last release in the 0.3.x series.
- Typo fix in README (#196, @MichaelSp).
- Bugfix for
InfluxDB::Client#list_series
when no series available (#195, @skladd). - Clarified/expanded docs (also #190, @paneq).
- Added preliminary
show_field_keys
method toInfluxDB::Client
(note: the API for this is not stable yet). - Degraded dependency on "cause" from runtime to development.
- Added option
discard_write_errors
to silently ignore errors when writing to the server (#182, @mickey). - Added
#list_series
and#delete_series
toInfluxDB::Client
(#183-186, @wrcola).
- You can now
InfluxDB::Client#query
,#write_points
,#write_point
and#write
now accept an additional parameter to override the database on invokation time (#173, #176, @jfragoulis).
- Bugfix for broken Unicode support (regression introduced in #169). Please note, this is only properly tested on Ruby 2.1+ (#171).
- Bugfix/Enhancement in
PointValue#escape
. Input strings are now scrubbed of invalid UTF byte sequences (#169, @ton31337).
- Bugfix in
Query::Builder#quote
(#168, @cthulhu666).
- Changed retry behaviour slightly. When the server responds with an incomplete response, we now assume a major server-side problem (insufficient resources, e.g. out-of-memory) and cancel any retry attempts (#165, #166).
- Added support for named and positional query parameters (#160, @retorquere).
- Fixed
prefix
handling for#ping
and#version
(#157, @dimiii).
- Added feature for JSON streaming response, via
"chunk_size"
parameter (#155, @mhodson-qxbranch).
- Reintroduced full dependency on "cause" (for Ruby 1.9 compat).
- Extended
Client#create_database
and#delete_database
to fallback onconfig.database
(#153, #154, @anthonator).
- Added resample options to
Client#create_continuous_query
(#149). - Fixed resample options to be Ruby 1.9 compatible (#150, @SebastianCoetzee).
- Mentioned in README, that 0.3.x series is the last one to support Ruby 1.9.
- Added resample options to
Client#create_continuous_query
(#149).
- Added config option to authenticate without credentials (#146, @pmenglund).
- Fixed #130 (again). Integer values are now really written as Integers to InfluxDB.
- Write queries are now checked against 204 No Content responses, in accordance with the official documentation (#128).
- Async options are now configurabe (#107).
- Empty tag keys/values are now omitted (#124).
- Async writer now behaves when stopping the client (#73).
- Update development dependencies and started enforcing Rubocop styles.
- Added
InfluxDB::Client#version
, returning the server version (#117). - Fixed escaping issues (#119, #121, #135).
- Integer values are now written as Integer, not as Float value (#131).
- Return all result series when querying multiple selects (#134).
- Made host cycling thread safe (#136).
- Added
epoch
option to client constructor and write methods (#104). - Added
#list_user_grants
(#111),#grant_user_admin_privileges
(#112) and#alter_retention_policy
(#114) methods.
- Fixed issues with Async client (#101)
- Avoid usage of
gsub!
(#102)
- Fix double quote tags escaping (#98)
- Large library refactoring (#88, #90)
- Extract config from client
- Extract HTTP functionality to separate module
- Extract InfluxDB management functions to separate modules
- Add writer concept
- Refactor specs (add cases)
- Add 'denormalize' option to config
- Recognize SeriesNotFound error
- Update README
- Add Rubocop config
- Break support for Ruby < 2
- Added support for InfluxDB 0.9+ (#92)
- last version to support InfluxDB 0.8.x