[
https://issues.apache.org/jira/browse/STORM-2937?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16354961#comment-16354961
]
Erik Weathers edited comment on STORM-2937 at 2/7/18 4:51 AM:
--------------------------------------------------------------
[~kabhwan]: good news, we have a seemingly working backport. I'm doing a
couple more sanity checks, but I'll send a PR later tonight.
Your suggestions above were very helpful, as well as [~Srdo]'s suggestion
(copied below) that we received over email which was critical to overcoming the
weirdness with the KafkaTridentSpoutEmitter (holy generics batman!).
{quote}
This PR
https://github.com/apache/storm/pull/1995/files#diff-72647db30ffd6005dc01c4d1f75d2c68
made a breaking change to
IOpaquePartitionedTridentSpout, so we'll have to do the same on 1.0.x.
{quote}
The other big thing that blocked us for awhile was a weird error where building
storm-kafka-client was failing because it couldn't find storm-core objects that
definitely existed and had been compiled. That happened due to the newer
branch's storm-kafka-client having reference to
{{<scope>${provided.scope}<scope>}}, which wasn't defined in storm's base
pom.xml back in 1.0.x-branch.
Regarding the commits -- clarity in communication through our code & its
historical footprints is by far the most important thing in my view. Sometimes
having an intentionally non-compiling commit is *definitely* the right thing to
do. e.g., renaming a java class and thus a file. If you don't do it as 2
separate commits then a future code maintainer/reader will not be able to
easily see what you did. (You will have made at least 1 change if not multiple
to the file -- those changes really should be separately tracked from the
entire contents of the file being moved.)
was (Author: erikdw):
[~kabhwan]: good news, we have a seemingly working backport. I'm doing a
couple more sanity checks, but I'll send a PR later tonight.
Your suggestions above, as well as [~Srdo]'s suggestion (copied below) that we
received over email was critical to overcoming the weirdness with the
KafkaTridentSpoutEmitter (holy generics batman!).
{quote}
This PR
https://github.com/apache/storm/pull/1995/files#diff-72647db30ffd6005dc01c4d1f75d2c68
made a breaking change to
IOpaquePartitionedTridentSpout, so we'll have to do the same on 1.0.x.
{quote}
The other big thing that blocked us for awhile was a weird error where building
storm-kafka-client was failing because it couldn't find storm-core objects that
definitely existed and had been compiled. That happened due to the newer
branch's storm-kafka-client having reference to
{{<scope>${provided.scope}<scope>}}, which wasn't defined in storm's base
pom.xml back in 1.0.x-branch.
Regarding the commits -- clarity in communication through our code & its
historical footprints is by far the most important thing in my view. Sometimes
having an intentionally non-compiling commit is *definitely* the right thing to
do. e.g., renaming a java class and thus a file. If you don't do it as 2
separate commits then a future code maintainer/reader will not be able to
easily see what you did. (You will have made at least 1 change if not multiple
to the file -- those changes really should be separately tracked from the
entire contents of the file being moved.)
> Overwrite storm-kafka-client 1.x-branch into 1.0.x-branch
> ---------------------------------------------------------
>
> Key: STORM-2937
> URL: https://issues.apache.org/jira/browse/STORM-2937
> Project: Apache Storm
> Issue Type: Task
> Components: storm-kafka-client
> Affects Versions: 1.0.6
> Reporter: Jungtaek Lim
> Assignee: Erik Weathers
> Priority: Blocker
>
> This is to track the effort of syncing up divergence between
> storm-kafka-client 1.x-branch and 1.0.x-branch so that critical fixes can be
> go in 1.0.x-branch as well.
> Note that it can modify storm-core as well (unlikely in a
> backwards-incompatible way but not 100% sure), so we should make a decision
> whether we allow the change in bugfix version line.
> Linking discussion thread:
> [https://lists.apache.org/thread.html/0451fed132bb982b618d9e0780282a87554f1bc5747827599f276944@%3Cdev.storm.apache.org%3E]
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
