Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
2.0.1
-
None
Description
When MaxScale is run under heavy load, it occasionally logs the following error message.
Could not find master among the backend servers. Previous master's state : Master is not in use
|
This can happen if a network disconnection happens before a query from the client has finished processing.
Here is an excerpt from a log created with a modified version of 2.0.1 with extra logging and session identifiers added to the log messages.
2016-10-16 14:34:08 notice : 1972: Closing bref: server1
|
2016-10-16 14:34:08 error : 1972 Backend hangup error handling.
|
2016-10-16 14:34:08 notice : 1972: Closing bref: server3
|
2016-10-16 14:34:08 error : 1972 Backend hangup error handling.
|
2016-10-16 14:34:08 notice : 1972: Closing bref: server2
|
2016-10-16 14:34:08 error : 1972 Backend hangup error handling.
|
2016-10-16 14:34:08 notice : 1972: Closing bref: server4
|
2016-10-16 14:34:08 error : 1972 Backend hangup error handling.
|
2016-10-16 14:34:08 error : 1972 Could not find master among the backend servers. Previous master's state : Master is not in use
|
This error message should not be logged if the master server was deliberately taken out of use.
Attachments
Issue Links
- relates to
-
-
- Closed
-