Thanks a lot Erick, Shawn and Walter. As Shawn mentioned I should not use
core admin API when in solrcloud, Just reloading collection solved the
problem.
--
Sent from: http://lucene.472066.n3.nabble.com/Solr-User-f472068.html
Core admin is required for REQUESTRECOVERY. Otherwise, I agree, don’t use it.
wunder
Walter Underwood
wun...@wunderwood.org
http://observer.wunderwood.org/ (my blog)
> On Apr 28, 2019, at 4:14 PM, Erick Erickson wrote:
>
> What Shawn said. Stop using the core admin API. It will not register th
What Shawn said. Stop using the core admin API. It will not register the core
with ZooKeeper. legacyCloud=true may put the info in ZooKeeper, but usually
puts it in clusterstate.json rather than the proper place,
collections>>your_collection>>state.json. There’s no guarantee whatsoever that
leg
On 4/28/2019 6:39 AM, ramyogi wrote:
Thanks Eric, After we create a collection and copy the index from one place
new place, we are doing UNLOAD core and CREATE core as below, is it wrong
and we have alternative to do that ?
Do not use CoreAdmin when running SolrCloud. At all. It will cause yo
Thanks Eric, After we create a collection and copy the index from one place
new place, we are doing UNLOAD core and CREATE core as below, is it wrong
and we have alternative to do that ?
r = requests.get('{}/admin/cores'.format(base_url), params={
'action': 'UNLOAD',
'core': co
There’s some documentation in the ref guide, did you see it?
Basically, if a replica is discovered by a solr instance by finding a
“core.properties” file under SOLR_HOME, legacyCloud=true will try to
reconstruct the ZooKeeper entry if it’s not there. This lead to lots of issues,
and so it has b
What is legacyCloud properties does. Is there a document I could refer. Why
it impacts collection creation with rule based and index copy from one to
another. Replicas not registered how to fix without legacyCloud=true
settings.
--
Sent from: http://lucene.472066.n3.nabble.com/Solr-User-f472068.
Do you have something about legacyCloud in your CLUSTERSTATUS response?
I have "properties":{"legacyCloud":"false"}
In the legacy cloud mode, also calles format 1, the state is stored in a
central clusterstate.js node in ZK, which does not scale well. In the
modern mode every collection has its
I found the reason,
&legacyCloud=true when I create a collection with this parameter I could
find that replicas data in CLUSTERSTATUS api request,. is there anything
wrong if I use this in SOLR 7.5.0 when create a collection ?
Please advice.
--
Sent from: http://lucene.472066.n3.nabble.com/Solr-
SOLR 7.5.0
Created collection as below:
/admin/collections?action=CREATE&name=test_collection&numShards=1&maxShardsPerNode=1&rule=replica:<2,node:*&collection.configName=test_collection_config
Created Successfully.
After that when I try to see CLUSTERSTATUS, it is giving "replicas": {}
empty.
10 matches
Mail list logo
