nicktelford commented on PR #15137:
URL: https://github.com/apache/kafka/pull/15137#issuecomment-1961419902
Closing this PR for now while I break changes out into separate PRs
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub an
nicktelford closed pull request #15137: KAFKA-14412: Transactional StateStores
URL: https://github.com/apache/kafka/pull/15137
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
T
nicktelford commented on PR #15137:
URL: https://github.com/apache/kafka/pull/15137#issuecomment-1910399080
> * Is it possible to have a separate PR for the config changes?
@cadonna Yes, in fact I'm working on that as I speak. :grin: It'll be a
little strange to essentially add config
cadonna commented on PR #15137:
URL: https://github.com/apache/kafka/pull/15137#issuecomment-1909913075
Some suggestions/questions to split the PR in smaller parts (without having
looked at the PR in much depth):
- Is it possible to have a separate PR for the config changes?
- Is it po
nicktelford commented on PR #15137:
URL: https://github.com/apache/kafka/pull/15137#issuecomment-1909852487
> Hey @nicktelford , I'll try to take a look at this here and there and I
can't speak for the Confluent folks, but I have to say it's pretty intimidating
to see a PR with 3000 LOC and
ableegoldman commented on PR #15137:
URL: https://github.com/apache/kafka/pull/15137#issuecomment-1909050890
Hey @nicktelford , I'll try to take a look at this here and there and I
can't speak for the Confluent folks, but I have to say it's pretty intimidating
to see a PR with 3000 LOC and
nicktelford commented on PR #15137:
URL: https://github.com/apache/kafka/pull/15137#issuecomment-1888900920
The build says there are "19 new failed tests", but they all pass locally.
Since all but 3 are in non-streams modules, this looks like poor CI performance
induced timeouts.
--
This
nicktelford commented on code in PR #15137:
URL: https://github.com/apache/kafka/pull/15137#discussion_r1443144457
##
streams/src/main/java/org/apache/kafka/streams/state/internals/RocksDBStore.java:
##
@@ -805,6 +959,194 @@ public void reset() {
public void close() {
nicktelford commented on code in PR #15137:
URL: https://github.com/apache/kafka/pull/15137#discussion_r1443141089
##
streams/src/main/java/org/apache/kafka/streams/processor/internals/ProcessorStateManager.java:
##
@@ -266,29 +269,20 @@ void initializeStoreOffsetsFromCheckpoint
nicktelford commented on code in PR #15137:
URL: https://github.com/apache/kafka/pull/15137#discussion_r1443138486
##
checkstyle/checkstyle.xml:
##
@@ -135,7 +135,7 @@
-
+
Review Comment:
`TaskExecutor#commitOffsetsOrTransaction`
--
This i
nicktelford commented on code in PR #15137:
URL: https://github.com/apache/kafka/pull/15137#discussion_r1443136941
##
streams/src/main/java/org/apache/kafka/streams/processor/internals/TaskManager.java:
##
@@ -262,8 +268,7 @@ private void closeDirtyAndRevive(final Collection
ta
nicktelford commented on PR #15137:
URL: https://github.com/apache/kafka/pull/15137#issuecomment-1878999651
@cadonna @mjsax @ableegoldman @lucasbru @wcarlson5 @bbejeck @vvcephei
@guozhangwang
My apologies to whoever draws the short straw of having to review this!
This is the m
nicktelford opened a new pull request, #15137:
URL: https://github.com/apache/kafka/pull/15137
See [KIP-892: Transactional Semantics for StateStores][1] for a detailed
explanation of the design.
---
Writes to StateStores are now isolated from the underlying RocksDB database
un
13 matches
Mail list logo
