incubator-cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jonathan Ellis <jbel...@gmail.com>
Subject Re: Encountering timeout exception when running get_key_range
Date Wed, 21 Oct 2009 20:52:10 GMT
On Wed, Oct 21, 2009 at 3:38 PM, Ramzi Rabah <rrabah@playdom.com> wrote:
> Hi Jonathan in what scenario will there be only one version of a row?

For instance, if you

insert columns A B C in row K
delete A B C from K

while row K is in memory, then when that memtable is flushed to an
sstable there will be one version of row K, the tombstone version.  if
you never update K again that will be the only version visible to
compaction.

> and for this scenario, does this mean that these tombstone records
> will never ever be deleted?

Yes. (before the fix attached to #507 anyway)

> Also on a higher level, what I am trying to do is provide some garbage
> collection of entries in the cassandra hash that have expired.(Kind of
> like time-to-live functionality for records in Cassandra). Do you have
> any insights on how that can be accomplished.

I don't understand what you mean by "gc of entries in the cassandra hash."

-Jonathan

Mime
View raw message