eG Help

Solr replication uses the master-slave model to distribute complete copies of a master index to one or more slave servers.The master server receives all updates and all changes are made against a single master server. Changes made on the master are distributed to all the slave servers which service all query requests from the clients. This enables Solr to remain responsive even with high query traffic.

By periodically scanning the replication process, the SolrReplicatTest keeps track of the requests and if there are any errors either on client or server side, and promptly notifies the administrators.

Using this test, administrators can effectively track the replication process and can handle if there are any errors in timely manner. This test reports the errors encountered during replication, time taken through the process and if there was any timeout.

Outputs of the test: One set of results for each Solr core in Apache Solr setup.

Measurement	Description	Measurement Unit	Interpretation
errors	Indicates the number of errors encountered by handler.	Number	Ideally the value of this measure should be 0.
clientErrors	Indicates the number of syntax or parse errors made by the client while making the request.	Number	Although can't do much about the client errors but you can take a look at errors reported just to ensure that they are informative enough.
serverErrors	Indicates the number of errors reported by the server while executing the request.	Number	The server errors should be investigated if they are because of missing documents or if because one of the components is malfunctioning.
requests	Indicates the number of requests of any kind made by client.	Number	It should be noted that all the requests are being replicated across the slave servers.
timeouts	Indicates the number of times when the response was received with partial results.	Number	Again it needs t be investigated if timeout happened because of lack of compute resources or cache unavailability etc.
handlerStart	Indicates the time it took for replication handler to be registered and started.	Minutes	The request handler for replication should start immediately as soon as data is updated on Solr server, if the delay is more than tolerance limit, you may need to investigate the same.