cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Eric Stevens <>
Subject Re: Storing PDF data on Cassandra db
Date Thu, 15 Jan 2015 17:35:55 GMT
@DENIZ, Jon's point is that CQL is the new standard, Thrift is frozen and
being deprecated.  Anything you build using the Thrift interface will hurt
you over time, so you ought to just go for CQL.  There really is next to no
reason not to use CQL aside from personal preference, and that argument
doesn't hold up well to defend an interface that is going away.  CQL is
already mature in 2.0 and above.  Read Robert Coli's excellent article on
what version of Cassandra you should be running:

As for storing PDFs in Cassandra, it's possible (you probably just want the
blob data type), but usually when people start talking about storing files
in Cassandra, it's because they're trying to do something Cassandra isn't a
great use case for.  If you want a distributed filesystem, you should look
for tools designed to support that.

However, if your files are relatively small, you probably won't run into
any real difficulty.  But if they start being megabytes or multiple
megabytes in size, and you use your Cassandra cluster for any other more
traditional purposes, you'll probably find that performance suffers for the
traditional purposes.

On Wed, Jan 14, 2015 at 1:18 AM, DENIZ <> wrote:

>  Hi ,
> I have checked The latest release is 2.1.2
> .Do you know when will The 3.0 release be available?
> Thanks in advance
>  ------------------------------
> Kimden: Jonathan Haddad <>
> Gönderme tarihi: ‎13.‎1.‎2015 16:26
> Kime:
> Konu: Re: Storing PDF data on Cassandra db
> For a new user, there's no point in learning Thrift if that user intends
> on upgrading past the version that they start with.  Thrift is a deprecated
> protocol and there's no new functionality going into it.  In 3.0 the
> sstable format is being upgraded to work primarily with native CQL
> partitions / rows and have a translation layer to work with thrift for
> compatibility.
> If you're simply storing k->v there's no reason (or advantage) to thrift
> over CQL.
> CREATE TABLE pdf ( pdf_id uuid primary key, content blob );
> will work just fine.  Adding additional fields as the system changes is
> convenient and works with any tool that reads CQL and won't require custom
> deserialization or any other nonsense.
> Jon
> On Tue Jan 13 2015 at 6:14:10 AM Peter Lin <> wrote:
>> you want to store the raw bytes, so look at examples for saving raw bytes.
>> I generally recommend using Thrift if you're going to do a lot of
>> read/write of binary data. CQL is good for primitive types, and maps/lists
>> of primitive types. I'm bias, but it's simpler and easier to use thrift for
>> storing file attachments.
>> On Tue, Jan 13, 2015 at 8:58 AM, DENIZ <> wrote:
>>>  Hi All,
>>> I want to store PDF documents on Cassandra db.What is the best way to
>>> store this type of data on Cassandra.How can I insert and select PDF file
>>> to the database.If possible can you explain with sample CQL statements
>>> Thanks in Advance
>>> Nil

View raw message