Cluster solr request for node state

Question

I am trying to set up our system health check to see the state of Solr nodes in our SolrCloud. I encountered the following problems:

We send a request to each of the Solr nodes separately. If we get a response and the status of the response is 0, we will assume that node is running. Unfortunately, we saw cases when the node is restored or even reduced, and the requests are still being processed.

Hoping to prevent this, we added a check that sends a ping request to solr. If the status returned by this request reads "OK", we assume that the node is inserted. Unfortunately, even with this request, if the node is restored or lowered, this check will not be completed.

My question is: What is the correct way to check node status in SolrCloud?

Answer 1

If you use SolrCloud, it is recommended that you support the explicit zookeeper ensemble . Because the zookeeper ensemble maintains the current SolrCloud state of each node and each splinter. This status is actually reflected from the SolrCloud administration window.

• Go to the admin window. Click "Cloud."
• Then click Tree to get the tree structure of SolrCloud architecture.
• Click /clusterstate.jsonto view SolrCloud status.

(clusterstate.json) json SolrCloud. , zookeeper, , SolrCloud.

• "zookeeper/installation/directory/bin"
• ./zkCli.sh -server ZK_IP:ZK_PORT (./zkCli.sh -server localhost: 2181)
• get /clusterstate.json

SolrCloud.

: ZK_IP - IP- HOST, . ZK_PORT - Zookeeper.

Answer 2

/clusterstate.json - , . ZooKeeper /live_nodes

Zookeeper , Solr Solr, , , , . Solr cloud, , , .

/live_nodes Solr node, , .

/live_nodes..., clusterstatus Solr .

http://localhost:8983/solr/admin/collections?action=clusterstatus&wt=json

clusterstatus Solr, , .. , .

SOLR_HOST solr.in.sh (/etc/default/solr.in.sh) - "localhost" - solr node. , Solr node, ZooKeeper "localhost" Solr node.