Jedis cluster pipeline sync problem

[rahul-narayana]

I am experimenting with the new cluster pipeline support in jedis. Here is a sample script in scala

val jedisClientConfig: DefaultJedisClientConfig = DefaultJedisClientConfig.builder().build()
val clusterConnectionProvider = new ClusterConnectionProvider(
      Set(
          new HostAndPort("h1", 1234),
          new HostAndPort("h2", 5678)
      ).asJava, jedisClientConfig)
var pipeline = new ClusterPipeline(clusterConnectionProvider)

As I am not using any pool I guess it is taking the default pool of size 8. I am syncing the pipeline after 10 commands.

pipeline.sync()
pipeline = new ClusterPipeline(clusterConnectionProvider)

After 80 commands, I guess the pool is exhausted and code freezes to move further. What is the best approach to handle such scenarios. Do I need to use new ClusterPipeline after every sync instead of using a ClusterConnectionProvider.

new ClusterPipeline approach makes a new connection for every sync which will slow down overall performance (running cluster slots again and again).

Please let me know what would be the ideal approach here. Thanks

[sazzad16]

close() ClusterPipeline after sync(). It'll release the resources (connections) it was using.

pipeline.sync()
pipeline.close()
pipeline = new ClusterPipeline(clusterConnectionProvider)

[rahul-narayana]

Thank you, It worked. The format is different between unclustered and clustered version so I missed this step to close the pipeline