[
https://issues.apache.org/jira/browse/CASSANDRA-1073?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12870978#action_12870978
]
Matthew F. Dennis commented on CASSANDRA-1073:
----------------------------------------------
Just to be clear, the test I automated:
parent process starts Cassandra server
parent process starts python thrift client in separate process
client loops inserting data
parent process sends SIGKILL to client in the middle of looping
parent process sends SIGTERM to server (is this not a clean shutdown?)
parent process starts server again
parent process connects to server over thrift from python
parent process issues get_range_slice and receives TimedOutException
As I understood Frank, he believed the issue to be that CliMain didn't call
disconnect because it was killed from Eclipse. When he explicitly typed "exit"
in the CLI, CliMain called disconnect and the problem did not present himself
(even though presumably he restarted Cassandra in the same way).
Should I be looking at killing the server instead?
> EOFException with Cassandra.Client.get_range_slices() API
> ---------------------------------------------------------
>
> Key: CASSANDRA-1073
> URL: https://issues.apache.org/jira/browse/CASSANDRA-1073
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Affects Versions: 0.6
> Reporter: Frank Du
> Assignee: Matthew F. Dennis
> Fix For: 0.6.3
>
>
> Below is the snippet I tried to run. The keyspace is named 'Keyspac1', with
> only 1 column family named 'CF1'.
> ColumnParent cp = new ColumnParent("CF1");
>
> SlicePredicate predicate = new SlicePredicate();
> SliceRange sliceRange = new SliceRange();
> sliceRange.setStart(new byte[0]);
> sliceRange.setFinish(new byte[0]);
> predicate.setSlice_range(sliceRange);
>
> KeyRange range = new KeyRange(10);
> range.setStart_key("".getBytes());
> range.setEnd_key("".getBytes());
>
> client.set_keyspace(keyspace);
> List<KeySlice> slices = client.get_range_slices(cp, predicate,
> range, ConsistencyLevel.ONE);
> Then an EOFException was spit out. For readability, it is included in the
> next comment.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
