How important is the name of the data collection?

[David Patterson <pa...@gmail.com>]

I'm trying to make two virtual machines, each with one 4.10 SOLR-cloud code
instance, connected to the same external Zookeeper site.

I want to create one data collection with one shard on each of these two
machines.

If I use the "start" method as described in the Apache Solr Reference Guide
for my release, will the two machines be connected if I declare the same
data collection name for both of them? If not, how do I connect them?

(I know the start method can make two solr-cloud instances on ONE virtual
machine, but I want to make one on each of two virtual machines.)

Thanks

Dave P

Re: How important is the name of the data collection?

[Erick Erickson <er...@gmail.com>]

OK, assuming that by "virtual machine" you're talking JVMs and not
VMWare or something....

bin/solr start -c -z localhost:2181 -p 8981 -s example/cloud/node1/solr
bin/solr start -c -z localhost:2181 -p 8982 -s example/cloud/node2/solr

Note that I cheat and use one of the example configs create
example/cloud/node1, example/cloud/node2 etc., then kill the Solr,
then start them individually as above.

It works just fine to create node1/solr and node2/solr somewhere, and
then just copy solr.xml and zoo.cfg into both

HTH,
Erick

On Mon, Jun 8, 2015 at 10:04 AM, David Patterson <pa...@gmail.com> wrote:
> I'm trying to make two virtual machines, each with one 4.10 SOLR-cloud code
> instance, connected to the same external Zookeeper site.
>
> I want to create one data collection with one shard on each of these two
> machines.
>
> If I use the "start" method as described in the Apache Solr Reference Guide
> for my release, will the two machines be connected if I declare the same
> data collection name for both of them? If not, how do I connect them?
>
> (I know the start method can make two solr-cloud instances on ONE virtual
> machine, but I want to make one on each of two virtual machines.)
>
> Thanks
>
> Dave P