how data is distributed?

[Suman Ganta]

I'm new to Hazelcast. I would like to know the scalability aspects of
Hazelcast. I understood that it implements distributed versions of the
collection APIs. When a key-value pair is pushed to a map, which node
actually holds that info? Will it be propagated to all nodes by
default? If so, if construct a map of 20k size all nodes of cluster
would contain that 20k map in memory? Or that 20k is distributed among
all nodes?

Regards,
-Suman.

[Talip Ozturk]

Everything is partitioned in Hazelcast so your map of 20k entries will
be partitioned over the cluster members.
if you have 10 node cluster then each node will hold 2k entries plus
2k backup of previous node. so in total each node will hold 4k entries
but of course each node will be able to access(read/write) all 20k
entries as if they were local.

for more info http://code.google.com/docreader/#p=hazelcast&s=hazelcast&t=InternalsDistributedMap

-talip

> --
>
> You received this message because you are subscribed to the Google Groups "Hazelcast" group.
> To post to this group, send email to haze...@googlegroups.com.
> To unsubscribe from this group, send email to hazelcast+...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/hazelcast?hl=en.
>
>
>