Definitely +1 to point 2) raised by Dawid. I'm not sure on points 1) and 3).
1) I can see the benefit of that but in reality most timestamp assigners
will probably need to be Serializable. If you look at my (updated) POC
branch [1] you can see how a TimestampAssigner would be specified on the
WatermarkStrategies helper class: [2]. The signature of this would have
to be changed to something like:
public <TA extends TimestampAssigner<T> & Serializable>
WatermarkStrategies<T> withTimestampAssigner(TA timestampAssigner)
Then, however, it would not be possible for users to specify a lambda or
anonymous inner function for the TimestampAssigner like this:
WatermarkStrategy<Long> testWmStrategy = WatermarkStrategies
.forGenerator(new PeriodicTestWatermarkGenerator())
.withTimestampAssigner((event, timestamp) -> event)
.build();
3) This makes sense if we only allow WatermarkStrategies on sources,
where the previous timestamp really is the "native" timestamp.
Currently, we also allow setting watermark strategies at arbitrary
points in the graph. I'm thinking we probably should only allow that in
sources but it's not the reality currently. I'm not against renaming it,
just voicing those thoughts.
Best,
Aljoscha
[1] https://github.com/aljoscha/flink/tree/flink-xxx-wm-generators-rebased
[2]
https://github.com/aljoscha/flink/blob/flink-xxx-wm-generators-rebased/flink-core/src/main/java/org/apache/flink/api/common/eventtime/WatermarkStrategies.java#L81
On 12.05.20 15:48, Stephan Ewen wrote:
+1 to all of Dawid's suggestions, makes a lot of sense to me
On Tue, May 12, 2020 at 2:32 PM Dawid Wysakowicz <dwysakow...@apache.org>
wrote:
Hi Aljoscha,
Sorry for adding comments during the vote, but I have some really minor
suggestions that should not influence the voting thread imo.
1) Does it make sense to have the TimestampAssigner extend from Flink's
Function? This implies it has to be serializable which with the factory
pattern is not strictly necessary, right? BTW I really like that you
suggested the FunctionInterface annotation there.
2) Could we rename the IdentityTimestampAssigner to e.g.
RecordTimestampAssigner/SystemTimestampAssigner/NativeTimestampAssigner...
Personally I found the IdentityTimestampAssigner a bit misleading as it
usually mean a no-op. Which did not click for me, as I assumed it
somehow returns the incoming record itself.
3) Could we rename the second parameter of TimestampAssigner#extract to
e.g. recordTimestamp/nativeTimestamp? This is similar to the point
above. This parameter was also a bit confusing for me as I thought at
times its somehow related to
TimerService#currentProcessingTimestamp()/currentWatermark() as the
whole system currentTimestamp.
Other than those three points I like the proposal and I was about to
vote +1 if it was not for those three points.
Best,
Dawid
On 11/05/2020 16:57, Jark Wu wrote:
Thanks for the explanation. I like the fatory pattern to make the member
variables immutable and final.
So +1 to the proposal.
Best,
Jark
On Mon, 11 May 2020 at 22:01, Stephan Ewen <se...@apache.org> wrote:
I am fine with that.
Much of the principles seem agreed upon. I understand the need to
support
code-generated extractors and we should support most of it already (as
Aljoscha mentioned via the factories) can extend this if needed.
I think that the factory approach supports code-generated extractors in
a
cleaner way even than an extractor with an open/init method.
On Mon, May 11, 2020 at 3:38 PM Aljoscha Krettek <aljos...@apache.org>
wrote:
We're slightly running out of time. I would propose we vote on the
basic
principle and remain open to later additions. This feature is quite
important to make the new Kafka Source that is developed as part of
FLIP-27 useful. Otherwise we would have to use the legacy interfaces in
the newly added connector.
I know that's a bit unorthodox but would everyone be OK with what's
currently there and then we iterate?
Best,
Aljoscha
On 11.05.20 13:57, Aljoscha Krettek wrote:
Ah, I meant to write this in my previous email, sorry about that.
The WatermarkStrategy, which is basically a factory for a
WatermarkGenerator is the replacement for the open() method. This is
the
same strategy that was followed for StreamOperatorFactory, which was
introduced to allow code generation in the Table API [1]. If we need
metrics or other things we would add that as a parameter to the
factory
method. What do you think?
Best,
Aljoscha
[1] https://issues.apache.org/jira/browse/FLINK-11974
On 10.05.20 05:07, Jark Wu wrote:
Hi,
Regarding to the `open()/close()`, I think it's necessary for
Table&SQL to
compile the generated code.
In Table&SQL, the watermark strategy and event-timestamp is defined
using
SQL expressions, we will
translate and generate Java code for the expressions. If we have
`open()/close()`, we don't need lazy initialization.
Besides that, I can see a need to report some metrics, e.g. the
current
watermark, the dirty timestamps (null value), etc.
So I think a simple `open()/close()` with a context which can get
MetricGroup is nice and not complex for the first version.
Best,
Jark
On Sun, 10 May 2020 at 00:50, Stephan Ewen <se...@apache.org> wrote:
Thanks, Aljoscha, for picking this up.
I agree with the approach of doing the here proposed set of changes
for
now. It already makes things simpler and adds idleness support
everywhere.
Rich functions and state always add complexity, let's do this in a
next
step, if we have a really compelling case.
On Wed, Apr 29, 2020 at 7:24 PM Aljoscha Krettek <
aljos...@apache.org>
wrote:
Regarding the WatermarkGenerator (WG) interface itself. The
proposal
is
basically to turn emitting into a "flatMap", we give the
WatermarkGenerator a "collector" (the WatermarkOutput) and the WG
can
decide whether to output a watermark or not and can also mark the
output
as idle. Changing the interface to return a Watermark (as the
previous
watermark assigner interface did) would not allow that flexibility.
Regarding checkpointing the watermark and keeping track of the
minimum
watermark, this would be the responsibility of the framework (or
the
KafkaConsumer in the current implementation). The user-supplied WG
does
not need to make sure the watermark doesn't regress.
Regarding making the WG a "rich function", I can see the potential
benefit but I also see a lot of pitfalls. For example, how should
the
watermark state be handled in the case of scale-in? It could be
made
to
work in the Kafka case by attaching the state to the partition
state
that we keep, but then we have potential backwards compatibility
problems also for the WM state. Does the WG usually need to keep
the
state or might it be enough if the state is transient, i.e. if you
have
a restart the WG would loose its histogram but it would rebuild it
quickly and you would get back to the same steady state as before.
Best,
Aljoscha
On 27.04.20 12:12, David Anderson wrote:
Overall I like this proposal; thanks for bringing it forward,
Aljoscha.
I also like the idea of making the Watermark generator a rich
function
--
this should make it more straightforward to implement smarter
watermark
generators. Eg, one that uses state to keep statistics about the
actual
out-of-orderness, and uses those statistics to implement a
variable
delay.
David
On Mon, Apr 27, 2020 at 11:44 AM Kostas Kloudas <
kklou...@gmail.com>
wrote:
Hi Aljoscha,
Thanks for opening the discussion!
I have two comments on the FLIP:
1) we could add lifecycle methods to the Generator, i.e. open()/
close(), probably with a Context as argument: I have not fully
thought
this through but I think that this is more aligned with the rest
of
our rich functions. In addition, it will allow, for example, to
initialize the Watermark value, if we decide to checkpoint the
watermark (see [1]) (I also do not know if Table/SQL needs to do
anything in the open()).
2) aligned with the above, and with the case where we want to
checkpoint the watermark in mind, I am wondering about how we
could
implement this in the future. In the FLIP, it is proposed to
expose
the WatermarkOutput in the methods of the WatermarkGenerator.
Given
that there is the implicit contract that watermarks are
non-decreasing, the WatermarkOutput#emitWatermark() will have (I
assume) a check that will compare the last emitted WM against the
provided one, and emit it only if it is >=. If not, then we risk
having the user shooting himself on the foot if he/she
accidentally
forgets the check. Given that the WatermarkGenerator and its
caller do
not know if the watermark was finally emitted or not (the
WatermarkOutput#emitWatermark returns void), who will be
responsible
for checkpointing the WM?
Given this, why not having the methods as:
public interface WatermarkGenerator<T> {
Watermark onEvent(T event, long eventTimestamp,
WatermarkOutput
output);
Watermark onPeriodicEmit(WatermarkOutput output);
}
and the caller will be the one enforcing any invariants, such as
non-decreasing watermarks. In this way, the caller can checkpoint
anything that is needed as it will have complete knowledge as to
if
the WM was emitted or not.
What do you think?
Cheers,
Kostas
[1] https://issues.apache.org/jira/browse/FLINK-5601
On Tue, Apr 21, 2020 at 2:25 PM Timo Walther <twal...@apache.org
wrote:
Thanks for the proposal Aljoscha. This is a very useful
unification.
We
have considered this FLIP already in the interfaces for FLIP-95
[1]
and
look forward to update to the new unified watermark generators
once
FLIP-126 has been accepted.
Regards,
Timo
[1] https://github.com/apache/flink/pull/11692
On 20.04.20 18:10, Aljoscha Krettek wrote:
Hi Everyone!
We would like to start a discussion on "FLIP-126: Unify (and
separate)
Watermark Assigners" [1]. This work was started by Stephan in
an
experimental branch. I expanded on that work to provide a PoC
for
the
changes proposed in this FLIP: [2].
Currently, we have two different flavours of Watermark
Assigners: AssignerWithPunctuatedWatermarks
and AssignerWithPeriodicWatermarks. Both of them extend
from TimestampAssigner. This means that sources that want to
support
watermark assignment/extraction in the source need to support
two
separate interfaces, we have two operator implementations for
the
different flavours. Also, this makes features such as generic
support
for idleness detection more complicated to implemented because
we
again
have to support two types of watermark assigners.
In this FLIP we propose two things:
Unify the Watermark Assigners into one Interface
WatermarkGenerator
Separate this new interface from the TimestampAssigner
The motivation for the first is to simplify future
implementations
and
code duplication. The motivation for the second point is again
code
deduplication, most assigners currently have to extend from
some
base
timestamp extractor or duplicate the extraction logic, or users
have
to
override an abstract method of the watermark assigner to
provide
the
timestamp extraction logic.
Additionally, we propose to add a generic wrapping
WatermarkGenerator
that provides idleness detection, i.e. it can mark a
stream/partition
as
idle if no data arrives after a configured timeout.
The "unify and separate" part refers to the fact that we want
to
unify
punctuated and periodic assigners but at the same time split
the
timestamp assigner from the watermark generator.
Please find more details in the FLIP [1]. Looking forward to
your feedback.
Best,
Aljoscha
[1]
https://cwiki.apache.org/confluence/display/FLINK/FLIP-126%3A+Unify+%28and+separate%29+Watermark+Assigners
[2] https://github.com/aljoscha/flink/tree/stephan-event-time
