take a look at:
https://issues.apache.org/jira/browse/SPARK-15798
On Dec 19, 2016 00:17, "Robin East" wrote:
This is also a feature we need for our time-series processing
> On 19 Dec 2016, at 04:07, Liang-Chi Hsieh wrote:
>
>
> Hi,
>
> As I know,
We will be interested by the results if you give a try to Dynamic allocation
with mesos !
- Mail Original -
De: "Michael Gummelt"
À: "Sumit Chawla"
Cc: u...@mesos.apache.org, d...@mesos.apache.org, "User"
,
The vote passed with the following +1 and -1:
+1
Reynold Xin*
Sean Owen*
Dongjoon Hyun
Xiao Li
Herman van Hövell tot Westerflier
Joseph Bradley*
Liwei Lin
Denny Lee
Holden Karau
Adam Roberts
vaquar khan
0/+1 (not sure what this means but putting it here just in case)
Felix Cheung
-1
Franklyn
Tim,
We will try to run the application in coarse grain mode, and share the
findings with you.
Regards
Sumit Chawla
On Mon, Dec 19, 2016 at 3:11 PM, Timothy Chen wrote:
> Dynamic allocation works with Coarse grain mode only, we wasn't aware
> a need for Fine grain mode
Dynamic allocation works with Coarse grain mode only, we wasn't aware
a need for Fine grain mode after we enabled dynamic allocation support
on the coarse grain mode.
What's the reason you're running fine grain mode instead of coarse
grain + dynamic allocation?
Tim
On Mon, Dec 19, 2016 at 2:45
Hey Prashant. Thanks for your codes. I did some investigation and it turned
out that ContextCleaner is too slow and its "referenceQueue" keeps growing.
My hunch is cleaning broadcast is very slow since it's a blocking call.
On Mon, Dec 19, 2016 at 12:50 PM, Shixiong(Ryan) Zhu <
> Is this problem of idle executors sticking around solved in Dynamic
Resource Allocation? Is there some timeout after which Idle executors can
just shutdown and cleanup its resources.
Yes, that's exactly what dynamic allocation does. But again I have no idea
what the state of dynamic
Great. Makes much better sense now. What will be reason to have
spark.mesos.mesosExecutor.cores more than 1, as this number doesn't include
the number of cores for tasks.
So in my case it seems like 30 CPUs are allocated to executors. And there
are 48 tasks so 48 + 30 = 78 CPUs. And i am
That makes sense. From the documentation it looks like the executors are
not supposed to terminate:
http://spark.apache.org/docs/latest/running-on-mesos.html#fine-grained-deprecated
> Note that while Spark tasks in fine-grained will relinquish cores as they
> terminate, they will not relinquish
Hey, Prashant. Could you track the GC root of byte arrays in the heap?
On Sat, Dec 17, 2016 at 10:04 PM, Prashant Sharma
wrote:
> Furthermore, I ran the same thing with 26 GB as the memory, which would
> mean 1.3GB per thread of memory. My jmap
>
> I should preassume that No of executors should be less than number of
tasks.
No. Each executor runs 0 or more tasks.
Each executor consumes 1 CPU, and each task running on that executor
consumes another CPU. You can customize this via
spark.mesos.mesosExecutor.cores (
I am using Apache Spark 2.0.2 and facing following issue while using
cartesian product in Spark Streaming module.
I am using compression codec as snappy but facing the same issue while using
the default one:LZ4, also using kryo for serialization.
I also see ample memory available in the executor
I think that what you are looking for is Dynamic resource allocation:
http://spark.apache.org/docs/latest/job-scheduling.html#dynamic-resource-allocation
Spark provides a mechanism to dynamically adjust the resources your application
occupies based on the workload. This means that your
Hi Chawla,
One possible reason is that Mesos fine grain mode also takes up cores
to run the executor per host, so if you have 20 agents running Fine
grained executor it will take up 20 cores while it's still running.
Tim
On Fri, Dec 16, 2016 at 8:41 AM, Chawla,Sumit
Yea, the idea is to use dynamic allocation. I can't speak to how well it
works with Mesos, though.
On Mon, Dec 19, 2016 at 11:01 AM, Mehdi Meziane
wrote:
> I think that what you are looking for is Dynamic resource allocation:
>
Since it’s not a regression from 2.0 (I believe the same issue affects both
2.0 and 2.1) it doesn’t merit a -1 vote according to the voting guidelines.
Of course, it would be nice if we could fix the various optimizer issues
that all seem to have a workaround that involves persist() (another one
But coarse grained does the exact same thing which i am trying to avert
here. At the cost of lower startup, it keeps the resources reserved till
the entire duration of the job.
Regards
Sumit Chawla
On Mon, Dec 19, 2016 at 10:06 AM, Michael Gummelt
wrote:
> Hi
>
> I
Hi
I don't have a lot of experience with the fine-grained scheduler. It's
deprecated and fairly old now. CPUs should be relinquished as tasks
complete, so I'm not sure why you're seeing what you're seeing. There have
been a few discussions on the spark list regarding deprecating the
-1 https://issues.apache.org/jira/browse/SPARK-18589 hasn't been resolved
by this release and is a blocker in our adoption of spark 2.0. I've updated
the issue with some steps to reproduce the error.
On Mon, Dec 19, 2016 at 4:37 AM, Sean Owen wrote:
> PS, here are the open
Hi all,
I perform sampling on a DStream by taking samples from RDDs in the DStream.
I have used two sampling mechanisms: simple random sampling and stratified
sampling.
Simple random sampling: inputStream.transform(x => x.sample(false,
fraction)).
Stratified sampling: inputStream.transform(x =>
PS, here are the open issues for 2.1.0. Forgot this one. No Blockers, but
one "Critical":
SPARK-16845
org.apache.spark.sql.catalyst.expressions.GeneratedClass$SpecificOrdering"
grows beyond 64 KB
SPARK-18669 Update Apache docs regard watermarking in Structured Streaming
SPARK-18894 Event time
This is also a feature we need for our time-series processing
> On 19 Dec 2016, at 04:07, Liang-Chi Hsieh wrote:
>
>
> Hi,
>
> As I know, Spark SQL doesn't provide native support for this feature now.
> After searching, I found only few database systems support it, e.g.,
22 matches
Mail list logo