0

I am trying to read rows from dev center and client apis but if i "select *" for all 2k rows it shows no host available exception in query trace though it has all the nodes up in cluster. but if i keep partition key in where condition it doesn't show any error. Can anyone help in finding the issue? Many thanks.

john cena
  • 41
  • 4

1 Answers1

2

You're probably hitting timeouts. Doing select * from with no where clause is not a good idea as you'll potentially be hitting every node in your cluster. If this is just for testing, you can bump up your read timeouts (read_request_timeout_in_ms) in the yaml but you don't want to use a query like that in production.

For further insight use the query tracing feature in Dev Center which is pretty nice in the latest releases.

phact
  • 7,305
  • 23
  • 27
  • Thanks for the reply @phact , the table just has 2k rows with just one partition key so i was thinking it should not overload the cluster. its basically a test i did.. Actual issue we had is we got many errors in client saying "nohostavailableexception" .. i see all nodes up and running.. so confused on what the issue can be.. i checked mem and cpu load all looks normal to me.. any information on this will be very helpful. thanks. – john cena Jun 17 '15 at 14:57