You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@spark.apache.org by James Yu <ja...@ispot.tv> on 2021/02/03 18:54:44 UTC
Poor performance caused by coalesce to 1
Hi Team,
We are running into this poor performance issue and seeking your suggestion on how to improve it:
We have a particular dataset which we aggregate from other datasets and like to write out to one single file (because it is small enough). We found that after a series of transformations (GROUP BYs, FLATMAPs), we coalesced the final RDD to 1 partition before writing it out, and this coalesce degrade the performance, not that this additional coalesce operation took additional runtime, but it somehow dictates the partitions to use in the upstream transformations.
We hope there is a simple and useful way to solve this kind of issue which we believe is quite common for many people.
Thanks
James
Re: Poor performance caused by coalesce to 1
Posted by Stéphane Verlet <st...@verlet.name>.
I had that issue too and from what I gathered, it is an expected optimization... Try using repartiion instead
Get BlueMail for Android
On Feb 3, 2021, 11:55, at 11:55, James Yu <ja...@ispot.tv> wrote:
>Hi Team,
>
>We are running into this poor performance issue and seeking your
>suggestion on how to improve it:
>
>We have a particular dataset which we aggregate from other datasets and
>like to write out to one single file (because it is small enough). We
>found that after a series of transformations (GROUP BYs, FLATMAPs), we
>coalesced the final RDD to 1 partition before writing it out, and this
>coalesce degrade the performance, not that this additional coalesce
>operation took additional runtime, but it somehow dictates the
>partitions to use in the upstream transformations.
>
>We hope there is a simple and useful way to solve this kind of issue
>which we believe is quite common for many people.
>
>
>Thanks
>
>James
Re: Poor performance caused by coalesce to 1
Posted by Gourav Sengupta <go...@gmail.com>.
Hi,
as always, I would like to first identify the problem before solving the
problem.
So to isolate the problem, first without coalesce try to write the data out
to a storage location and check the time.
Then try to do coalesce to one and check the time.
If the time between writing down between coalesce and writing out to the
files is very large, then the issue is coalesce. Otherwise the issue is the
chain of transformations before coalesce.
Anyways, its 2021, and I always get confused when people use RDD's. Any
particular reason why dataframes would not work?
Regards,
Gourav Sengupta
On Wed, Feb 3, 2021 at 7:20 PM James Yu <ja...@ispot.tv> wrote:
> Hi Silvio,
>
> The result file is less than 50 MB in size so I think it is small and
> acceptable enough for one task to write.
>
> Your suggestion sounds interesting. Could you guide us further on how to
> easily "add a stage boundary"?
>
> Thanks
> ------------------------------
> *From:* Silvio Fiorito <si...@granturing.com>
> *Sent:* Wednesday, February 3, 2021 11:05 AM
> *To:* James Yu <ja...@ispot.tv>; user <us...@spark.apache.org>
> *Subject:* Re: Poor performance caused by coalesce to 1
>
>
> Coalesce is reducing the parallelization of your last stage, in your case
> to 1 task. So, it’s natural it will give poor performance especially with
> large data. If you absolutely need a single file output, you can instead
> add a stage boundary and use repartition(1). This will give your query full
> parallelization during processing while at the end giving you a single task
> that writes data out. Note that if the file is large (e.g. in 1GB or more)
> you’ll probably still notice slowness while writing. You may want to
> reconsider the 1-file requirement for larger datasets.
>
>
>
> *From: *James Yu <ja...@ispot.tv>
> *Date: *Wednesday, February 3, 2021 at 1:54 PM
> *To: *user <us...@spark.apache.org>
> *Subject: *Poor performance caused by coalesce to 1
>
>
>
> Hi Team,
>
>
>
> We are running into this poor performance issue and seeking your
> suggestion on how to improve it:
>
>
>
> We have a particular dataset which we aggregate from other datasets and
> like to write out to one single file (because it is small enough). We
> found that after a series of transformations (GROUP BYs, FLATMAPs), we
> coalesced the final RDD to 1 partition before writing it out, and this
> coalesce degrade the performance, not that this additional coalesce
> operation took additional runtime, but it somehow dictates the partitions
> to use in the upstream transformations.
>
>
>
> We hope there is a simple and useful way to solve this kind of issue which
> we believe is quite common for many people.
>
>
>
>
>
> Thanks
>
>
>
> James
>
Re: Poor performance caused by coalesce to 1
Posted by Silvio Fiorito <si...@granturing.com>.
As I suggested, you need to use repartition(1) in place of coalesce(1)
That will give you a single file output without losing parallelization for the rest of the job.
From: James Yu <ja...@ispot.tv>
Date: Wednesday, February 3, 2021 at 2:19 PM
To: Silvio Fiorito <si...@granturing.com>, user <us...@spark.apache.org>
Subject: Re: Poor performance caused by coalesce to 1
Hi Silvio,
The result file is less than 50 MB in size so I think it is small and acceptable enough for one task to write.
Your suggestion sounds interesting. Could you guide us further on how to easily "add a stage boundary"?
Thanks
________________________________
From: Silvio Fiorito <si...@granturing.com>
Sent: Wednesday, February 3, 2021 11:05 AM
To: James Yu <ja...@ispot.tv>; user <us...@spark.apache.org>
Subject: Re: Poor performance caused by coalesce to 1
Coalesce is reducing the parallelization of your last stage, in your case to 1 task. So, it’s natural it will give poor performance especially with large data. If you absolutely need a single file output, you can instead add a stage boundary and use repartition(1). This will give your query full parallelization during processing while at the end giving you a single task that writes data out. Note that if the file is large (e.g. in 1GB or more) you’ll probably still notice slowness while writing. You may want to reconsider the 1-file requirement for larger datasets.
From: James Yu <ja...@ispot.tv>
Date: Wednesday, February 3, 2021 at 1:54 PM
To: user <us...@spark.apache.org>
Subject: Poor performance caused by coalesce to 1
Hi Team,
We are running into this poor performance issue and seeking your suggestion on how to improve it:
We have a particular dataset which we aggregate from other datasets and like to write out to one single file (because it is small enough). We found that after a series of transformations (GROUP BYs, FLATMAPs), we coalesced the final RDD to 1 partition before writing it out, and this coalesce degrade the performance, not that this additional coalesce operation took additional runtime, but it somehow dictates the partitions to use in the upstream transformations.
We hope there is a simple and useful way to solve this kind of issue which we believe is quite common for many people.
Thanks
James
Re: Poor performance caused by coalesce to 1
Posted by James Yu <ja...@ispot.tv>.
Hi Silvio,
The result file is less than 50 MB in size so I think it is small and acceptable enough for one task to write.
Your suggestion sounds interesting. Could you guide us further on how to easily "add a stage boundary"?
Thanks
________________________________
From: Silvio Fiorito <si...@granturing.com>
Sent: Wednesday, February 3, 2021 11:05 AM
To: James Yu <ja...@ispot.tv>; user <us...@spark.apache.org>
Subject: Re: Poor performance caused by coalesce to 1
Coalesce is reducing the parallelization of your last stage, in your case to 1 task. So, it’s natural it will give poor performance especially with large data. If you absolutely need a single file output, you can instead add a stage boundary and use repartition(1). This will give your query full parallelization during processing while at the end giving you a single task that writes data out. Note that if the file is large (e.g. in 1GB or more) you’ll probably still notice slowness while writing. You may want to reconsider the 1-file requirement for larger datasets.
From: James Yu <ja...@ispot.tv>
Date: Wednesday, February 3, 2021 at 1:54 PM
To: user <us...@spark.apache.org>
Subject: Poor performance caused by coalesce to 1
Hi Team,
We are running into this poor performance issue and seeking your suggestion on how to improve it:
We have a particular dataset which we aggregate from other datasets and like to write out to one single file (because it is small enough). We found that after a series of transformations (GROUP BYs, FLATMAPs), we coalesced the final RDD to 1 partition before writing it out, and this coalesce degrade the performance, not that this additional coalesce operation took additional runtime, but it somehow dictates the partitions to use in the upstream transformations.
We hope there is a simple and useful way to solve this kind of issue which we believe is quite common for many people.
Thanks
James
Re: Poor performance caused by coalesce to 1
Posted by Silvio Fiorito <si...@granturing.com>.
Coalesce is reducing the parallelization of your last stage, in your case to 1 task. So, it’s natural it will give poor performance especially with large data. If you absolutely need a single file output, you can instead add a stage boundary and use repartition(1). This will give your query full parallelization during processing while at the end giving you a single task that writes data out. Note that if the file is large (e.g. in 1GB or more) you’ll probably still notice slowness while writing. You may want to reconsider the 1-file requirement for larger datasets.
From: James Yu <ja...@ispot.tv>
Date: Wednesday, February 3, 2021 at 1:54 PM
To: user <us...@spark.apache.org>
Subject: Poor performance caused by coalesce to 1
Hi Team,
We are running into this poor performance issue and seeking your suggestion on how to improve it:
We have a particular dataset which we aggregate from other datasets and like to write out to one single file (because it is small enough). We found that after a series of transformations (GROUP BYs, FLATMAPs), we coalesced the final RDD to 1 partition before writing it out, and this coalesce degrade the performance, not that this additional coalesce operation took additional runtime, but it somehow dictates the partitions to use in the upstream transformations.
We hope there is a simple and useful way to solve this kind of issue which we believe is quite common for many people.
Thanks
James
Re: Poor performance caused by coalesce to 1
Posted by Mich Talebzadeh <mi...@gmail.com>.
That sounds like a plan as suggested by Sean, I have also seen caching the
RS before coalesce provides benefits, especially for a minute 50MB data.
Check Spark GUI storage tab for its effect.
HTH
Mich
LinkedIn * https://www.linkedin.com/profile/view?id=AAEAAAAWh2gBxianrbJd6zP6AcPCCdOABUrV8Pw
<https://www.linkedin.com/profile/view?id=AAEAAAAWh2gBxianrbJd6zP6AcPCCdOABUrV8Pw>*
*Disclaimer:* Use it at your own risk. Any and all responsibility for any
loss, damage or destruction of data or any other property which may arise
from relying on this email's technical content is explicitly disclaimed.
The author will in no case be liable for any monetary damages arising from
such loss, damage or destruction.
On Wed, 3 Feb 2021 at 19:08, Sean Owen <sr...@gmail.com> wrote:
> Probably could also be because that coalesce can cause some upstream
> transformations to also have parallelism of 1. I think (?) an OK solution
> is to cache the result, then coalesce and write. Or combine the files after
> the fact. or do what Silvio said.
>
> On Wed, Feb 3, 2021 at 12:55 PM James Yu <ja...@ispot.tv> wrote:
>
>> Hi Team,
>>
>> We are running into this poor performance issue and seeking your
>> suggestion on how to improve it:
>>
>> We have a particular dataset which we aggregate from other datasets and
>> like to write out to one single file (because it is small enough). We
>> found that after a series of transformations (GROUP BYs, FLATMAPs), we
>> coalesced the final RDD to 1 partition before writing it out, and this
>> coalesce degrade the performance, not that this additional coalesce
>> operation took additional runtime, but it somehow dictates the partitions
>> to use in the upstream transformations.
>>
>> We hope there is a simple and useful way to solve this kind of issue
>> which we believe is quite common for many people.
>>
>>
>> Thanks
>>
>> James
>>
>
Re: Poor performance caused by coalesce to 1
Posted by Sean Owen <sr...@gmail.com>.
Probably could also be because that coalesce can cause some upstream
transformations to also have parallelism of 1. I think (?) an OK solution
is to cache the result, then coalesce and write. Or combine the files after
the fact. or do what Silvio said.
On Wed, Feb 3, 2021 at 12:55 PM James Yu <ja...@ispot.tv> wrote:
> Hi Team,
>
> We are running into this poor performance issue and seeking your
> suggestion on how to improve it:
>
> We have a particular dataset which we aggregate from other datasets and
> like to write out to one single file (because it is small enough). We
> found that after a series of transformations (GROUP BYs, FLATMAPs), we
> coalesced the final RDD to 1 partition before writing it out, and this
> coalesce degrade the performance, not that this additional coalesce
> operation took additional runtime, but it somehow dictates the partitions
> to use in the upstream transformations.
>
> We hope there is a simple and useful way to solve this kind of issue which
> we believe is quite common for many people.
>
>
> Thanks
>
> James
>