I have a cluster with 18 task managers 4 task slots each running a
job whose source/sink(s) are declared with FlinkSQL using the Kafka
connector. The topic being read has 36 partitions. The problem I'm
observing is that the subtasks for the sources are not evenly
distributed. For example, 1 tas
Hi Aeden,
the option that you mentioned should have actually caused your desired
behavior. Can you double-check that it's set for the job (you can look at
the config in the Flink UI to be 100% sure).
Another option is to simply give all task managers 2 slots. In that way,
the scheduler can only e
Hi Arvid,
Thanks for responding. I did check the configuration tab of the job
manager and the setting cluster.evenly-spread-out-slots: true is
there. However I'm still observing unevenness in the distribution of
source tasks. Perhaps this additional information could shed light.
Version: 1.12.1
Hi Aeden,
just to be sure: All task managers have the same hardware/memory
configuration, haven't they? I'm not 100% sure whether this affects the
slot selection in the end, but it looks like this parameter has also an
influence on the slot matching strategy preferring slots with less
utilization o
Hi Matthias,
Yes, all the task managers have the same hardware/memory configuration.
Aeden
On Fri, Mar 12, 2021 at 3:25 AM Matthias Pohl wrote:
>
> Hi Aeden,
> just to be sure: All task managers have the same hardware/memory
> configuration, haven't they? I'm not 100% sure whether this affects
Is this a brand-new job, with the cluster having all 18 TMs at the time
of submission? (or did you add more TMs while the job was running)
On 3/12/2021 5:47 PM, Aeden Jameson wrote:
Hi Matthias,
Yes, all the task managers have the same hardware/memory configuration.
Aeden
On Fri, Mar 12, 202
Hi Aeden,
IIUC, the topic being read has 36 partitions means that your source task
has a parallelism of 36. What's the parallelism of other tasks? Is the job
taking use of all the 72 (18 TMs * 4 slots/TM) slots?
I'm afraid currently there's no good way to guarantee subtasks of a task
are spread o
Hi Xintong,
Thanks for replying. Yes, you understood my scenario. Every task
has the same parallelism since we're using FlinkSql unless there is a
way to change the parallelism of the source task that I have missed.
Your explanation of the setting makes sense and is what I ended up
concluding
If all the tasks have the same parallelism 36, your job should only
allocate 36 slots. The evenly-spread-out-slots option should help in your
case.
Is it possible for you to share the complete jobmanager logs?
Thank you~
Xintong Song
On Tue, Mar 16, 2021 at 12:46 AM Aeden Jameson
wrote:
>
There may be a slight misunderstanding: all the FlinkSql tasks _were_
set at a parallelism of 72 -- 18 nodes 4 slots. I was hoping that the
setting cluster.evenly-spread-out-slots would spread out the active
kafka consumers evenly among the TM's given the topic has 36
partitions, but I now realize
10 matches
Mail list logo