You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@iceberg.apache.org by Jaguar Xiong <xi...@gmail.com> on 2019/09/14 12:10:38 UTC
concurrent write
Hi,
New joiner here!
I just find this project while reading about delta-lake and apache hudi. And this project seems very unique with its "table format".
Is there some further info about "concurrent write" part? So far I could only find a few paragraph. Especially what's the general idea of handing write conflict? Or is this more targeting on the "append only" type of tables.
Thanks!
Jaguar
Re: concurrent write
Posted by Ryan Blue <rb...@netflix.com.INVALID>.
Looks like the attachment was removed. Here's a link to the presentation:
http://iceberg.apache.org/Concurrency%20in%20Iceberg.pdf
On Thu, Sep 19, 2019 at 12:22 PM Ryan Blue <rb...@netflix.com> wrote:
> Hi Jaguar,
>
> Thanks for your interest! I'm attaching the slides for a talk I gave at
> Adobe that covered how concurrency works in Iceberg. Hopefully the examples
> give you an idea about how it works. We can build all of the usual database
> operations using this framework, so it isn't limited to just appends. This
> flexibility is one of the reasons why we think Iceberg is a good candidate
> to be an interchange standard for at-rest data between big data projects
> like Spark, Presto, Hudi, NiFi, etc.
>
> rb
>
> On Sat, Sep 14, 2019 at 5:10 AM Jaguar Xiong <xi...@gmail.com>
> wrote:
>
>> Hi,
>> New joiner here!
>> I just find this project while reading about delta-lake and apache
>> hudi. And this project seems very unique with its "table format".
>> Is there some further info about "concurrent write" part? So far I
>> could only find a few paragraph. Especially what's the general idea of
>> handing write conflict? Or is this more targeting on the "append only" type
>> of tables.
>>
>> Thanks!
>> Jaguar
>>
>
>
> --
> Ryan Blue
> Software Engineer
> Netflix
>
--
Ryan Blue
Software Engineer
Netflix
Re: concurrent write
Posted by Ryan Blue <rb...@netflix.com.INVALID>.
Hi Jaguar,
Thanks for your interest! I'm attaching the slides for a talk I gave at
Adobe that covered how concurrency works in Iceberg. Hopefully the examples
give you an idea about how it works. We can build all of the usual database
operations using this framework, so it isn't limited to just appends. This
flexibility is one of the reasons why we think Iceberg is a good candidate
to be an interchange standard for at-rest data between big data projects
like Spark, Presto, Hudi, NiFi, etc.
rb
On Sat, Sep 14, 2019 at 5:10 AM Jaguar Xiong <xi...@gmail.com> wrote:
> Hi,
> New joiner here!
> I just find this project while reading about delta-lake and apache hudi.
> And this project seems very unique with its "table format".
> Is there some further info about "concurrent write" part? So far I could
> only find a few paragraph. Especially what's the general idea of handing
> write conflict? Or is this more targeting on the "append only" type of
> tables.
>
> Thanks!
> Jaguar
>
--
Ryan Blue
Software Engineer
Netflix