yep I meant 120 per second :)
On Fri, Mar 31, 2017 at 11:19 AM, Ted Yu wrote:
> The 1,2million seems to be European notation.
>
> You meant 1.2 million, right ?
>
> On Mar 31, 2017, at 1:19 AM, Kamil Dziublinski <
> kamil.dziublin...@gmail.com> wrote:
>
> Hi,
>
> Thanks
The 1,2million seems to be European notation.
You meant 1.2 million, right ?
> On Mar 31, 2017, at 1:19 AM, Kamil Dziublinski
> wrote:
>
> Hi,
>
> Thanks for the tip man. I tried playing with this.
> Was changing fetch.message.max.bytes (I still have 0.8 kafka)
Hi,
Thanks for the tip man. I tried playing with this.
Was changing fetch.message.max.bytes (I still have 0.8 kafka) and
also socket.receive.buffer.bytes. With some optimal settings I was able to
get to 1,2 million reads per second. So 50% increase.
But that unfortunately does not increase when I
I'm wondering what I can tweak further to increase this. I was reading in this
blog: https://data-artisans.com/extending-the-yahoo-streaming-benchmark/
about 3 millions per sec with only 20 partitions. So i'm sure I should be able
to squeeze out more out of it.
Not really sure if it is relevant
Thanks Ted, will read about it.
While we are on throughput.
Do you guys have any suggestion on how to optimise kafka reading from
flink?
In my current setup:
Flink is on 15 machines on yarn
Kafka on 9 brokers with 40 partitions. Source parallelism is 40 for flink,
And just for testing I left only
Kamil:
In the upcoming hbase 2.0 release, there are more write path optimizations
which would boost write performance further.
FYI
> On Mar 30, 2017, at 1:07 AM, Kamil Dziublinski
> wrote:
>
> Hey guys,
>
> Sorry for confusion it turned out that I had a bug in
Hey guys,
Sorry for confusion it turned out that I had a bug in my code, when I was
not clearing this list in my batch object on each apply call. Forgot it has
to be added since its different than fold.
Which led to so high throughput. When I fixed this I was back to 160k per
sec. I'm still
Hi Kamil,
the performance implications might be the result of which state the
underlying functions are using internally. WindowFunctions use ListState
or ReducingState, fold() uses FoldingState. It also depends on the size
of your state and the state backend you are using. I recommend the
Hi guys,
I’m using flink on production in Mapp. We recently swapped from storm.
Before I have put this live I was doing performance tests and I found
something that “feels” a bit off.
I have a simple streaming job reading from kafka, doing window for 3
seconds and then storing into hbase.