How to handle a down node in high Throughput

[deenbandhu-agarwal]

Hello,

I want to handle scenarios gracefully(without any error) when a node is going down because of some issue or we are restarting a node manually. I tried retry but it is not helping can somebody help me this?

[deenbandhu-agarwal]

Any update on this ?

[deenbandhu-agarwal]

@jhecking Any update on this?

[deenbandhu-agarwal]

Hi @jhecking

Any response to this. Its been a long time now.

[khaf]

Hi @deenbandhu-agarwal, did you manage to work around this issue? In most circumstances the client should not raise any exceptions. Can you provide a gist so that we can better understand your use case?

[mubashirkamran4]

Hi @khaf I am still getting the same error on batch_get method of client object. I get it very frequently. This is how I am using it.

      client = Client.new(config['servers'], policy: { timeout: 0.1 })
      batch = {}
      client.batch_get(
        keys,
        [],
        timeout: timeout,
        max_retries: 5,
        sleep_between_retries: 0.5,
        replica: Replica::SEQUENCE
      ).zip(ids) do |record, id|
        batch[id] = record.nil? ? {} : record.bins
      end```