0

I have a single node cluster and when I perform the following query, I get the following error, 

DELETE FROM issuetable WHERE "partitionId"='1';

**

  • Error:

** : 

WriteTimeout: Error from server: code=1100 [Coordinator node timed 
out waiting for replica nodes' responses] message="Operation timed out 
- received only 0 responses." info={'received_responses': 0, 
'required_responses': 1, 'consistency': 'QUORUM'}

As per datastax document (https://datastax.github.io/python-driver/api/cassandra.html), they mentioned increasing the write_request_timeout_in_ms in the cassandra.yaml

Question :

1) Why I get this error? As per my table (issuetable) I have only 84 entries under that partition.

2) Should this be solved client-side or server-side?

Vadim Kotov
  • 8,084
  • 8
  • 48
  • 62
Harry
  • 3,072
  • 6
  • 43
  • 100
  • whats your replication factor, and how many nodes/dcs do you have? Looks like you have a RF < 3 so if the 1 node that owns the data is in a GC or something it will timeout. – Chris Lohfink Oct 03 '18 at 14:43
  • @ChrisLohfink CREATE KEYSPACE issuekeyspace WITH replication = {'class': 'SimpleStrategy', 'replication_factor': '1'} AND durable_writes = true; so how to avoid this issue? Do I need to tune any parameter? – Harry Oct 03 '18 at 15:03
  • with a RF of one its same as using CL.ALL, that node has a GC and you will have timeouts. To have availability when a node is unavailable you need at least 3 replicas – Chris Lohfink Oct 03 '18 at 15:11

0 Answers0