cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Richard grossman <richie...@gmail.com>
Subject Re: Best Way to bulk delete rows
Date Fri, 06 Nov 2009 17:27:02 GMT
Dou you mean to delete the files physically on the disk for this CF after
the flush of course will be ok but will need a restart for each server in
the cluster ?


On Fri, Nov 6, 2009 at 7:18 PM, Michael Greene <michael.greene@gmail.com>wrote:

> `nodeprobe flush` is just to make sure you don't have data in the
> commitlog for the column families.  The deletion of the appropriate
> data files is the actual erasure.  As long as you only delete the db
> files for the column family you want to remove, you will be fine.
>
> Michael
>
> On Fri, Nov 6, 2009 at 11:14 AM, Richard grossman <richiesgr@gmail.com>
> wrote:
> > Thanks for fast answer,
> > The problem is that I need to remove only 1 column family in the keystore
> is
> > the nodeproble flush don't erase all the column families ?
> >
> > I'll enjoy to work on any task that can help the community and help me to
> > better understand cassandra
> >
> > Thanks
> >
> > On Fri, Nov 6, 2009 at 7:05 PM, Jonathan Ellis <jbellis@gmail.com>
> wrote:
> >>
> >> nodeprobe flush (requires 0.4.2), remove the data files, restart.
> >>
> >> it would be nice to make this more user-friendly (and not require
> >> restart): https://issues.apache.org/jira/browse/CASSANDRA-531
> >>
> >> if you'd like to work on this let me know, I can point you in the
> >> right direction.
> >>
> >> On Fri, Nov 6, 2009 at 10:55 AM, Richard grossman <richiesgr@gmail.com>
> >> wrote:
> >> > Hi
> >> >
> >> > I need to delete large amount of data at each update ~3000000 rows
> (the
> >> > entire ColumnFamily in fact) is there any solution to make faster than
> >> > just
> >> > calling the delete API.
> >> >
> >> > Thanks
> >> >
> >> > Richard
> >
> >
>

Mime
View raw message