Partition data - advantage and disadvantage

2009-12-28 Thread JKnight JKnight
Dear all, Could you show me the advantage and disadvantage of each partition type: ~ org.apache.cassandra.dht.RandomPartitioner, ~ org.apache.cassandra.dht.OrderPreservingPartitioner, and ~ org.apache.cassandra.dht.CollatingOrderPreservingPartitioner. Thank a lot for support. -- Best

Re: Partition data - advantage and disadvantage

2009-12-28 Thread Jonathan Ellis
On Mon, Dec 28, 2009 at 9:07 AM, Joe Stump j...@joestump.net wrote:    ~ org.apache.cassandra.dht.OrderPreservingPartitioner, and    ~ org.apache.cassandra.dht.CollatingOrderPreservingPartitioner. They're both ordering partitioners. The difference is that one is binary order and the other is

EasyCassandra.pm Perl interface alpha 0.01

2009-12-28 Thread Ted Zlatanov
Attached is the first alpha (0.01) version of my EasyCassandra.pm Perl interface to Cassandra. I am also attaching a demo script that will show the intended usage, but basically the idea is that the user can express gets, puts, and removals in shorthand like Subscribed/-1[] to mean the latest

Re: Partition data - advantage and disadvantage

2009-12-28 Thread Ted Zlatanov
On Mon, 28 Dec 2009 08:07:18 -0700 Joe Stump j...@joestump.net wrote: JS The advantage of the random partitioner is that it randomly JS distributes your keys across the cluster. This (theoretically) JS avoids key clustering on nodes. The big disadvantage is that you JS can't do key range

Re: Partition data - advantage and disadvantage

2009-12-28 Thread Ted Zlatanov
On Mon, 28 Dec 2009 09:53:56 -0700 Joe Stump j...@joestump.net wrote: JS On Dec 28, 2009, at 9:51 AM, Ted Zlatanov wrote: If each node does a key enumeration, can the results be aggregated somehow? It seems useful to get a list of all the keys across the cluster even if it's not 100%

Re: Partition data - advantage and disadvantage

2009-12-28 Thread Joe Stump
On Dec 28, 2009, at 11:00 AM, Ted Zlatanov wrote: Is this worth a JIRA feature request? Or is it something Cassandra will never support fully? From the user's perspective it's very useful. I don't know why it'd be very useful to be honest. Lots of us have CF's with billions of keys. Ours,

Re: Partition data - advantage and disadvantage

2009-12-28 Thread Ted Zlatanov
On Mon, 28 Dec 2009 11:02:30 -0700 Joe Stump j...@joestump.net wrote: JS On Dec 28, 2009, at 11:00 AM, Ted Zlatanov wrote: Is this worth a JIRA feature request? Or is it something Cassandra will never support fully? From the user's perspective it's very useful. JS I don't know why it'd be

Re: Partition data - advantage and disadvantage

2009-12-28 Thread Ted Zlatanov
On Mon, 28 Dec 2009 11:44:27 -0700 Joe Stump j...@joestump.net wrote: JS On Dec 28, 2009, at 11:40 AM, Ted Zlatanov wrote: I can see that's a problem. In my case, row keys represent switches in production so I don't expect more than a few hundred. An application can't find out how many