How to create replicas after data loaded

I have been trying to find a solution to this here on this group and many
Google searches.

I set up ES with 5 shards per index, 0 replicas. I create a new index by
day and hour and have loaded content. After loading 750M records, I end up
with 1425 shards across 285 indices. Everything is working fine.

Now I want to add a replica, so I use the curl command:

curl -XPUT 'localhost:9200/*/_settings' -d '{ "index": {
"number_of_replicas" : 1}}'

Which also worked fine - I can query ES and see that each index is marked
as having 1 replica. The shards however that would be the replica shards
are marked as Unassigned.

How do I go about forcing ES to actually replicate the indexes now into
these replica shards?

Thanks,
Jerry

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

On Tue, Nov 19, 2013 at 5:04 PM, Jerry Russell jerryrussell@gmail.comwrote:

Which also worked fine - I can query ES and see that each index is marked

as having 1 replica. The shards however that would be the replica shards
are marked as Unassigned.

How do I go about forcing ES to actually replicate the indexes now into
these replica shards?

My "is it plugged in" question: do you have more than one node in your
cluster? Like, check localhost:9200/_cluster/health and make sure.

Beyond that, does the health api tell you that some are being assigned?
Are there any logs about failure to assign?

Nik

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

How many nodes do you have?

--
David Pilato | Technical Advocate | Elasticsearch.com
@dadoonet | @elasticsearchfr

Le 19 novembre 2013 at 23:04:49, Jerry Russell (jerryrussell@gmail.com) a écrit:

I have been trying to find a solution to this here on this group and many Google searches.

I set up ES with 5 shards per index, 0 replicas. I create a new index by day and hour and have loaded content. After loading 750M records, I end up with 1425 shards across 285 indices. Everything is working fine.

Now I want to add a replica, so I use the curl command:

curl -XPUT 'localhost:9200/*/_settings' -d '{ "index": { "number_of_replicas" : 1}}'

Which also worked fine - I can query ES and see that each index is marked as having 1 replica. The shards however that would be the replica shards are marked as Unassigned.

How do I go about forcing ES to actually replicate the indexes now into these replica shards?

Thanks,
Jerry

You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

One node. Will replicas not be created if there is only one node?

Thanks,
Jerry

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

No. It does not really make sense to replicate data on the same node.

--
David Pilato | Technical Advocate | Elasticsearch.com
@dadoonet | @elasticsearchfr

Le 19 novembre 2013 at 23:13:27, Jerry Russell (jerryrussell@gmail.com) a écrit:

One node. Will replicas not be created if there is only one node?

Thanks,
Jerry

You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

ES will intentionally avoid replicating shards on the same node so the
cluster is always redundant. You'd need to add another node for it to
automatically replicate.

-Andreas

On Tuesday, November 19, 2013 2:13:24 PM UTC-8, Jerry Russell wrote:

One node. Will replicas not be created if there is only one node?

Thanks,
Jerry

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

I understand it doesn't make sense to replicate with one node. But I am
seeing a weird problem with searching that I was hoping was related to not
having a replica (always getting different number of search results for
same search with static data). This problem I am running into did not occur
on same data set when using two nodes with replicas, so I needed to confirm
in the current environment if replicas were the real cause or not.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.