I'd be interested in what the folks who want CAS implementations think about
vector clocks. Can you use them to fulfill your use cases? If not, why not?
I ask because I have found myself wanting CAS in Cassandra too, but I think
that's only because I'm pretty familiar with HTTP. I think vector clo
S>: An *atomic* CAS is another beast and I see at least two difficulties:
S>: 1) making it atomic locally: Cassandra's implementation is very much
>multi-threaded. On a given node, while you're reading-comparing-and-swapping
>on some column c, no other thread should be allowed to write c (even '
Hi,
I was looking a bit on a case we have with columnfamily which has 439k
supercolumns, each supercolumn with ~3 columns each so a total of some 1.3
million objects in total.
This takes about 9 second to read with thrift on first access, 4-5 second on
second access.
I took a little closer look
On Tue, Jun 8, 2010 at 9:44 PM, Anty wrote:
> In theory ,it could do some help.
I tried this and unfortunately there was no noticeable difference.
-Brandon
On Mon, Jun 21, 2010 at 11:19 PM, Rishi Bhardwaj wrote:
> I have read the post on cages and it is definitely very interesting. But
> cages seems to be too coarse grained compared to an Atomic Compare and Swap
> on Cassandra column value. Cages would makes sense when one wants to do
> multiple atom
