Hi,

I also have loads of tombstones while only inserting new rows without ever 
deleting rows.

My rows contain null columns and also collections.

How can I avoid the creation of these tombstones?

Thanks for your help.

On 24 Mar,2016, at 02:08, Steve Robenalt 
<sroben...@highwire.org<mailto:sroben...@highwire.org>> wrote:

Hi Henry,

Since new values are written without checking for previous values, the only way 
to assure that a nulled column blocks a previously valid one is to write a 
tombstone to occlude whatever value might have been present. I believe there 
are some recent changes as to the handling of null values to relax some of this 
behavior under some circumstances, but I'm not sure as to which version it 
affects or what the circumstances are. A similar question was raised recently 
on this list, so the list archives may be of some use.

Steve


On Wed, Mar 23, 2016 at 6:00 PM, Henry M 
<henrymanm...@gmail.com<mailto:henrymanm...@gmail.com>> wrote:
What is the reason for the tombstone for a brand new insert? Do the fields get 
written as a whole (both nulls and non-nulls?

I understand the rationale for tombstones for deletes and updates but it does 
not make sense for an insert (I am trying to make sense of it). I understand 
Cassandra writes the record without first checking its existence but wouldn't 
the whole set of fields including values to null out be applied as one single 
operation?

Thank you,
Henry



On Wed, Mar 23, 2016 at 5:47 PM, Eric Stevens 
<migh...@gmail.com<mailto:migh...@gmail.com>> wrote:
In addition to writing null values acting as tombstones, also INSERTing a 
collection (or UPDATE where you set the collection rather than append to it) 
are also operations which will create tombstones.

On Wed, Mar 23, 2016 at 12:09 PM Robert Coli 
<rc...@eventbrite.com<mailto:rc...@eventbrite.com>> wrote:
On Wed, Mar 23, 2016 at 9:50 AM, Ralf Steppacher 
<ralf.viva...@gmail.com<mailto:ralf.viva...@gmail.com>> wrote:
How come I end up with that large a number of tombstones?

Are you inserting NULLs?

=Rob





--
Steve Robenalt
Software Architect
sroben...@highwire.org<mailto:bza...@highwire.org>
(office/cell): 916-505-1785

HighWire Press, Inc.
425 Broadway St, Redwood City, CA 94063
www.highwire.org<http://www.highwire.org/>

Technology for Scholarly Communication

Reply via email to