8/22/2018 2:28:51 PM
3.51.4
(#1933)
Flow trigger classes will be serialized and persisted in quartz tables and deserialized back.
If serialVersionUID is not defined specifically, JVM would generate a default id in runtime which is dependent on compiler implementation, and and can thus result in unexpected nvalidClassExceptions during deserialization if new deployment has different serialVersionUID from old deployment which has been persisted in db.
Why choose those numbers(-1330280892166841227L ...)?
To be backward compatible since those are the serialVersionUID of first deployment of flow trigger classes and already have been persisted in database.
|
|
|
8/16/2018 3:48:32 PM
Fix flaky test and remove sleep time in FlowRunnerYamlTest.
* Conditional workflow - remove the support for condition on job props in phase one.
* Revert "Fix flaky test and remove sleep time in FlowRunnerYamlTest."
This reverts commit a2ac558ed5ca9af16f3dfe94ce0f9815ac7b6f37.
|
8/16/2018 1:39:31 PM
happens if the bar above (the one that has Submit User, Duration etc.) spans to multiple rows. Even one of them is enough to cause overflowing positioning, but the screenshot above demonstrates an extreme case where the tab is barely visible any more.
As a quick fix, I copied the same rules that are set for div#graphView.container-fill (you can find it right above the changed lines).
the fix sets a fixed height to the log textbox, and as the screenshot shows it gets cut off if the window size is not as high. There must be a way to make the position relative properly.. If that is found, why not apply that better rule also for div#graphView.container-fill?
|
8/16/2018 2:29:21 AM
unit tests for execution dispatch failures
Fix a minor bug in ExecutorManager: off by 1 in check of azkaban.maxDispatchingErrors:
reference.getNumErrors() > this.maxDispatchingErrors
->
reference.getNumErrors() >= this.maxDispatchingErrors
This makes sense to me: for example, if dispatching has failed 2 times and maxDispatchingErrors=2, don't try to dispatch for a 3rd time.
|
8/15/2018 6:00:17 PM
that upload API doesn't return an error when a deleted project zip is uploaded.
When a project is deleted, Azkaban just marks it to be inactive today. In this PR, we fix this issue by adding the check. 410 Gone client error response code indicates that access to the target resource is no longer available at the origin server and that this condition is likely to be permanent.
|
8/13/2018 6:38:59 PM
inside condition. (#1912)
|
8/10/2018 5:07:09 PM
with maven convention. (#1915)
move all java files into 'src/main/java'.
|
8/10/2018 3:01:07 PM
example setupExecutors() logs a similar message. However fetching active executors is quick.
Loading the active flows is what happens after it, and it may take a while (last time in my case it took ~20s). It's comforting to see in the log what we're waiting for, when the server is starting.
|
8/10/2018 2:56:40 PM
as the last statement of a loop doesn't change anything
|
