Commit graph

51 commits

Author SHA1 Message Date
Harlow Ward
0328cba5c9 adjust channel ownership for group 2019-06-02 17:46:28 -07:00
Harlow Ward
2ab5ec4031 revert examples 2019-06-02 17:26:24 -07:00
Harlow Ward
bd42663013 Introduce Group interface and AllGroup
As we move towards consumer groups we'll need to support the current
"consume all shards" strategy, and setup the codebase for the
anticipated "consume balanced shards."
2019-06-02 17:25:44 -07:00
James Greenhill
b48acfa5d4 Add Mysql support for checkpointing (#87) 2019-04-12 22:15:49 -07:00
Harlow Ward
97fe4e66ff
Use shard broker to monitor and process new shards (#85)
* Use shard broker to start processing new shards

The addition of a shard broker will allow the consumer to be notified
when new shards are added to the stream so it can consume them.

Fixes: https://github.com/harlow/kinesis-consumer/issues/36
2019-04-09 22:03:12 -07:00
Harlow Ward
c4f363a517 Example data is in repo, no need to download 2019-04-07 16:33:56 -07:00
Harlow Ward
76158d24ab
Introduce ScanFunc signature and remove ScanStatus (#77)
Major changes:

```go
type ScanFunc func(r *Record) error
```

* Simplify the callback func signature by removing `ScanStatus` 
* Leverage context for cancellation 
* Add custom error `SkipCheckpoint` for special cases when we don't want to checkpoint

Minor changes:

* Use kinesis package constants for shard iterator types
* Move optional config to new file

See conversation on #75 for more details
2019-04-07 16:29:12 -07:00
Harlow Ward
2f58b136fe Add dummy users data for producing onto stream
Pulled from: https://github.com/awslabs/amazon-kinesis-connectors
2018-12-30 07:28:18 -08:00
Harlow Ward
4f374e4425 Update example to use new AWS Session 2018-12-29 10:45:26 -08:00
Harlow Ward
94f0b2ae1e Fix import cycle error for postgres tests 2018-12-29 10:37:24 -08:00
Harlow Ward
fb98fbe244
Remove the client wrapper (#58)
Having an additional Client has added some confusion (https://github.com/harlow/kinesis-consumer/issues/45) on how to provide a
custom kinesis client. Allowing `WithClient` to accept a Kinesis client
it cleans up the interface.

Major changes:

* Remove the Client wrapper; prefer using kinesis client directly
* Change `ScanError` to `ScanStatus` as the return value isn't necessarily an error

Note: these are breaking changes, if you need last stable release please see here: https://github.com/harlow/kinesis-consumer/releases/tag/v0.2.0
2018-07-28 22:53:33 -07:00
Emanuel Ramos
b7be26418a Add postgres checkpoint implementation (#55) 2018-06-17 19:27:10 -07:00
Prometheus
739e9e39a5 Make it possible to let user use 3rd party logging library (#56) 2018-06-12 18:07:33 -07:00
Prometheus
e6a489c76b Scanerror signals the consumer if we should continue scanning for next record & whether to checkpoint. (#54)
* remove ValidateCheckpoint

* update for checkpoint can not customize retryer

* implement the scan error as in PR 44

* at least log if record processor has error

* mistakenly removed this line

* propage error up. ignore invalid state
2018-06-08 08:40:42 -07:00
Prometheus
d058203b6e Make what aws error to trigger retry decided by caller (#52)
* remove ValidateCheckpoint
* make retrying on error decided by caller
2018-06-04 20:07:58 -07:00
Prometheus
9a7e102a05 remove ValidateCheckpoint (#51) 2018-06-01 19:34:46 -07:00
Prometheus
992cc42419 DDB uses default AWS config settings to ping table; won't work with WithDyanmoClient. Misc update on example and README (#50) 2018-06-01 16:14:42 -07:00
Prometheus
9e0a97916d Use AWS resource iface, overwrite default dynamodb, more explicit in example about overwrite default AWS resrouce client (#49)
* use custom kinesis client

* use aws sdk interface, add missing api for ddb

* add overwrite default dynamodbclien usage
2018-05-31 17:41:14 -07:00
Anant Prakash
2a5856ec99 Correct AWS_REGION (#48) 2018-05-30 06:45:38 -07:00
Harlow Ward
64cdf69249
Add interval flush for DDB checkpoint (#40)
Add interval flush for DDB checkpoint

* Allow checkpointing on a specified interval
* Add shutdown method to checkpoint to force flush

Minor changes:

* Swap order of input params for checkpoint (app, table)

Addresses: https://github.com/harlow/kinesis-consumer/issues/39
2017-12-30 20:21:10 -08:00
Harlow Ward
058f383e30 Add cancellation of pipeline from signal interrupts 2017-11-26 16:00:03 -08:00
Harlow Ward
89570130f5 Leverage bigger batchsize when seeding example data 2017-11-26 15:59:17 -08:00
Harlow Ward
edf0467eb0 Format errors from caller 2017-11-23 11:29:58 -08:00
Harlow Ward
86f1df782e Return the shard scan errors to top-level caller 2017-11-23 08:49:37 -08:00
Harlow Ward
6401371727 Simplify checkpoint interface; reduce input vars 2017-11-22 20:01:31 -08:00
Harlow Ward
4ffe3ec55a Add logs for start scan and checkpoints 2017-11-22 17:52:41 -08:00
Harlow Ward
3770136f64 Allow user to override no-op checkpoint with Option 2017-11-22 17:44:42 -08:00
Harlow Ward
c91f6233ef Add counter for exposing scanner metrics 2017-11-22 14:10:11 -08:00
Harlow Ward
90d2903fe6 Use stdlib logging, default to discard 2017-11-22 10:46:39 -08:00
Harlow Ward
4d6a85e901 Make the Checkpoint a required input for Consumer
The Checkpoint functionality is an important part of the library and
previously it wasn't obvious that the Consumer was defaulting to Redis
for this functionality.

* Add Checkpoint as required param for new consumer
2017-11-21 08:58:16 -08:00
Harlow Ward
8d2cc5bc20 Return error from scan instead of terminating the program 2017-11-20 11:45:41 -08:00
Harlow Ward
6ee965ec0a
Add DDB as consumer checkpoint option (#37)
* Simplify the checkpoint interface
* Add DDB backend for checkpoint persistence

Implements: https://github.com/harlow/kinesis-consumer/issues/26
2017-11-20 09:37:30 -08:00
Harlow Ward
130c78456c
Simplify the consumer experience (#35)
Major changes:

* Remove intermediate batching of kinesis records
* Call the callback func with each record
* Use functional options for config 

https://dave.cheney.net/2014/10/17/functional-options-for-friendly-apis

Minor changes:

* update README messaging about Kinesis -> Firehose functionality
* remove unused buffer and emitter code
2017-11-20 08:21:40 -08:00
Harlow Ward
ceca88b96a Add required fields to Config
Taking some inspiration from:
https://github.com/tj/go-kinesis/blob/master/kinesis.go#L50-L75
2016-05-07 18:10:31 -07:00
Harlow Ward
f4c8d8090d Use TJ's Kinesis producer library 2016-05-01 16:54:47 -07:00
Harlow Ward
a12c15a191 Use no-op Info logging by default 2016-05-01 12:40:30 -07:00
Harlow Ward
afae1bea36 Use config object for optional params
After reading notes from Peter's talk I like the idea of using a config
object where consumers of the library can override the defaults.

https://peter.bourgon.org/go-best-practices-2016/#configuration
2016-05-01 12:20:44 -07:00
Harlow Ward
3aa0f72efe Add logging when records are emitted w/ record count 2016-05-01 10:43:42 -07:00
Harlow Ward
d5bdd3f4bc Leverage AWS S3 retries 2016-05-01 10:42:28 -07:00
Harlow Ward
7fccc9da9f Update function name to represent functionality 2016-04-30 09:14:25 -07:00
Harlow Ward
c29698550f Add config options to Consumer
The Firehose service can take a max batch size of 500. While created the
example the need for finer grained configuration was necessary.
2016-02-09 22:31:15 -08:00
Harlow Ward
aae268108e Move emitters to separate packages 2016-02-08 19:39:09 -08:00
Harlow Ward
caa2fc0b0a Add Kinesis Firehose example 2016-02-08 13:21:54 -08:00
Harlow Ward
f0e6461cb6 Refactor to use handler func
The previous pipeline model required a lot of setup and abstracted away
the processing of records. By passing a HandlerFunc to the consumer we
keep the business logic of processing of records closer to the use of
the consumer.

* Add refactoring note and SHA to README
2016-02-06 17:50:17 -08:00
Harlow Ward
509f68de89 Add error checking for tmp file 2016-01-09 09:52:47 -08:00
Harlow Ward
d542fa996f Use AWS SDK and concurrency 2016-01-09 09:46:28 -08:00
Harlow Ward
0d5e9b7b02 Use AWS SDK for S3 Emitter
* Handle retries for S3 Emitter
* Update Pipeline to use AWS backoff
* Leverage aws error kit in recoverable error helpers
2015-08-16 17:52:10 -07:00
Harlow Ward
18173842fb Use AWS SDK
Limit the amount of library dependencies by using the official AWS SDK.

https://github.com/harlow/kinesis-connectors/issues/19
2015-08-15 23:17:49 -07:00
Harlow Ward
3d9e6e2485 Add Redshift Pipeline example 2015-05-25 23:18:14 -07:00
Harlow Ward
cd54569c61 Use go-kit Log interface 2015-05-25 22:40:40 -07:00