4/12/2016 11:38:40 AM
Pierre-Alexandre Meyer <pierre@mouraf.org>
|
4/11/2016 10:04:26 AM
to be required to print stack traces, but not anymore: http://slf4j.org/faq.html#paramException
Signed-off-by: Pierre-Alexandre Meyer <pierre@mouraf.org>
|
4/10/2016 6:04:47 PM
messages by using key-value pairs whenever possible
(i.e. key1='value1', key2='value2') - this should help parsing in Splunk, Logstash, etc.
* Change default logback.xml pattern (use ISO8601 format for dates, switch to key/values, shorten logger name)
* Review log levels and use error only for unexpected errors (e.g. codepath that should never happen)
Signed-off-by: Pierre-Alexandre Meyer <pierre@mouraf.org>
|
|
|
4/16/2016 6:50:21 PM
approach to first record an attempt during priorCall and then mark it as success during the after call.
Add the logics (and test) to verify that if system were to crash after payment but before afterCall, the code would not end up
doing a double payment.
|
4/15/2016 8:16:56 PM
paymentCookieId) when notifying for new invoice payment. See #508
|
4/15/2016 6:57:46 PM
rows are in success state and ignore them otherwise
|
4/14/2016 9:49:48 PM
multiple initial transactions (AUTH, PURCHASE, CREDIT) for the same paymentId in either SUCCESS or PENDING state.
- Removed state machine transition AUTH_SUCCESS -> AUTH_INIT to ensure state machine prevent multiple transaction for when the satte was SUCCESS
- Added some logic in PaymentLeavingStateCallback to handle the PENDING use case
- Deleted test for #371 (which is limked to that ticket.
- Added a series of tests to verfify the behavior of creating new initial transactions for various TransactionStatus/Same or Different Keys
|
4/14/2016 6:58:19 PM
create a PENDING AUTH and then more status to SUCCESS during completion.
(Also use the same transactionExternalKey for auth completion)
|
|
|
