Yingyi Bu created ASTERIXDB-1492:
------------------------------------

             Summary: State mess up for failed Feeds
                 Key: ASTERIXDB-1492
                 URL: https://issues.apache.org/jira/browse/ASTERIXDB-1492
             Project: Apache AsterixDB
          Issue Type: Bug
            Reporter: Yingyi Bu
            Assignee: Abdullah Alamoudi


I had a failed feed because of timeout.  Then, I dropped the dataverse and 
created the same feed again.  After connecting the feed again, I got the 
following events in the NC log:

{noformat}
2016-06-21 14:20:34,603 WARN  [cb-io-1-15] dcp.DCPHandler 
(Slf4JLogger.java:warn(161)) - Unexpected status of StreamRequestResponse: 
ERR_EXISTS (0x2, The key exists in the cluster (with another CAS value))

2016-06-21 14:20:34,603 WARN  [cb-io-1-15] dcp.DCPHandler 
(Slf4JLogger.java:warn(161)) - Unexpected status of StreamRequestResponse: 
ERR_EXISTS (0x2, The key exists in the cluster (with another CAS value))

2016-06-21 14:20:34,604 WARN  [cb-io-1-15] dcp.DCPHandler 
(Slf4JLogger.java:warn(161)) - Unexpected status of StreamRequestResponse: 
ERR_EXISTS (0x2, The key exists in the cluster (with another CAS value))

2016-06-21 14:20:34,604 WARN  [cb-io-1-15] dcp.DCPHandler 
(Slf4JLogger.java:warn(161)) - Unexpected status of StreamRequestResponse: 
ERR_EXISTS (0x2, The key exists in the cluster (with another CAS value))

2016-06-21 14:20:34,604 WARN  [cb-io-1-15] dcp.DCPHandler 
(Slf4JLogger.java:warn(161)) - Unexpected status of StreamRequestResponse: 
ERR_EXISTS (0x2, The key exists in the cluster (with another CAS value))

2016-06-21 14:20:34,605 WARN  [cb-io-1-15] dcp.DCPHandler 
(Slf4JLogger.java:warn(161)) - Unexpected status of StreamRequestResponse: 
ERR_EXISTS (0x2, The key exists in the cluster (with another CAS value))

2016-06-21 14:20:34,605 WARN  [cb-io-1-15] dcp.DCPHandler 
(Slf4JLogger.java:warn(161)) - Unexpected status of StreamRequestResponse: 
ERR_EXISTS (0x2, The key exists in the cluster (with another CAS value))
{noformat}

After wiping out the AsterixDB instance and redoing everything from scratch,  
it works.  Therefore, I doubt that the system keeps some states around for a 
failed feed.





--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to