On a quick glance those look OK, what commands did you use _exactly_
to create your new collection? The names are a bit odd and it's not
clear how they could have gotten that way. how many documents have you
tried to index to your new collection? Any errors in the logs?
And how many documents are indexed to the single shard in your tests?
It's unlikely but possible that if you're using just a few docs they
all happen to hash into one shard.
bq: In the existing enviroment, i see only 1 core for all shards
Also not sure what's going on here. Did you perhaps use the _core_
admin API on the original collection? The Collections API imposes a
naming convention.
We need to know the exact steps you used in each case, exactly how
you're indexing documents etc.
Best,
Erick
On Mon, Dec 14, 2015 at 1:41 AM, ig01 <inna.gel...@elbitsystems.com> wrote:
> I have an existing solrcloud 4.4 configured with zookeeper.
> The current setting is 3 shards, each shard has a leader and replica. All
> are mapped to the same collection1.
>
> {"collection1":{
> "shards":{
> "shard1":{
> "range":"80000000-d554ffff",
> "state":"active",
> "replicas":{
> "core_node4":{
> "state":"active",
> "core":"collection1",
> "node_name":"10.200.101.132:9983_solr",
> "base_url":"http://10.200.101.132:9983/solr";,
> "leader":"true"},
> "core_node7":{
> "state":"active",
> "core":"collection1",
> "node_name":"10.200.101.131:8983_solr",
> "base_url":"http://10.200.101.131:8983/solr"}}},
> "shard2":{
> "range":"d5550000-2aa9ffff",
> "state":"active",
> "replicas":{
> "core_node2":{
> "state":"active",
> "core":"collection1",
> "node_name":"10.200.101.131:9983_solr",
> "base_url":"http://10.200.101.131:9983/solr"},
> "core_node5":{
> "state":"active",
> "core":"collection1",
> "node_name":"10.200.101.133:8983_solr",
> "base_url":"http://10.200.101.133:8983/solr";,
> "leader":"true"}}},
> "shard3":{
> "range":"2aaa0000-7fffffff",
> "state":"active",
> "replicas":{
> "core_node3":{
> "state":"active",
> "core":"collection1",
> "node_name":"10.200.101.132:8983_solr",
> "base_url":"http://10.200.101.132:8983/solr"},
> "core_node6":{
> "state":"active",
> "core":"collection1",
> "node_name":"10.200.101.133:9983_solr",
> "base_url":"http://10.200.101.133:9983/solr";,
> "leader":"true"}}}},
> "router":"compositeId"}}
>
>
>
> I have downloaded solrcloud 5.2.1 and ran solr.cmd. Created almost the same
> setting with 2 shards each shard has replica and leader.
>
> {"collection1":{
> "replicationFactor":"1",
> "shards":{
> "shard1_0":{
> "range":"80000000-ffffffff",
> "state":"active",
> "replicas":{
> "core_node3":{
> "core":"collection1_shard1_0_replica1",
> "base_url":"http://10.1.20.31:8983/solr";,
> "node_name":"10.1.20.31:8983_solr",
> "state":"active",
> "leader":"true"},
> "core_node5":{
> "core":"collection1_shard1_0_replica2",
> "base_url":"http://10.1.20.31:7574/solr";,
> "node_name":"10.1.20.31:7574_solr",
> "state":"active"}}},
> "shard1_1":{
> "range":"0-7fffffff",
> "state":"active",
> "replicas":{
> "core_node4":{
> "core":"collection1_shard1_1_replica1",
> "base_url":"http://10.1.20.31:8983/solr";,
> "node_name":"10.1.20.31:8983_solr",
> "state":"active",
> "leader":"true"},
> "core_node6":{
> "core":"collection1_shard1_1_replica2",
> "base_url":"http://10.1.20.31:7574/solr";,
> "node_name":"10.1.20.31:7574_solr",
> "state":"active"}}}},
> "router":{"name":"compositeId"},
> "maxShardsPerNode":"1",
> "autoAddReplicas":"false",
> "autoCreated":"true"}}
>
> The problem is when i am indexing a document to http://10.1.20.31:8983/solr,
> it is only indexed to collection1_shard1_0_replica1, it is not spreading the
> documents to the other shard.. why is that ? is the solr configured well ?
> In the existing enviroment, i see only 1 core for all shards, in the new
> there are 2 cores for each shard ?
>
> Please advise.
>
>
>
>
> --
> View this message in context:
> http://lucene.472066.n3.nabble.com/solr-cloud-invalid-shard-collection-configuration-tp4245151.html
> Sent from the Solr - User mailing list archive at Nabble.com.
