Details
-
Bug
-
Status: Open (View Workflow)
-
Major
-
Resolution: Unresolved
-
5.5.37-galera
-
CentOS6.2
three virtual PC, use 'birdge model' to connect network;
Description
Step:
1.Three nodes all run;
2.set session autocommit=0(In the first node);
3.call pp();--insert into a table 1000,000 rows,I also lifted the restrictions(In the first node);
4.commit at the first node,and after seconds and before it completion，I shutdown the node,but the other node'status is 'split-brain';
Attachments
Activity
| Field | Original Value | New Value |
|---|---|---|
| Labels | QFY_test_mariadb_galera_cluster |
| Assignee | Nirbhay Choubey [ nirbhay_c ] | |
| Labels | galera |
| Workflow | defaullt [ 40408 ] | MariaDB v2 [ 42993 ] |
| Workflow | MariaDB v2 [ 42993 ] | MariaDB v3 [ 62026 ] |
| Assignee | Nirbhay Choubey [ nirbhay_c ] |
| Assignee | Sachin Setiya [ sachin.setiya.007 ] |
| Fix Version/s | 5.5-galera [ 21900 ] |
| Fix Version/s | 5.5-galera [ 21900 ] | |
| Assignee | Sachin Setiya [ sachin.setiya.007 ] | Stepan Patryshev [ stepan.patryshev ] |
| Labels | galera | galera need_verification |
| Fix Version/s | N/A [ 14700 ] |
| Assignee | Stepan Patryshev [ stepan.patryshev ] | Ramesh Sivaraman [ JIRAUSER48189 ] |
| Workflow | MariaDB v3 [ 62026 ] | MariaDB v4 [ 139653 ] |
| Component/s | Galera [ 10124 ] |
If both 2nd and 3rd node were connected to the first node only (but not to each other), I'd think it's quite normal that the shutdown of the first node caused the split; although, maybe I'm missing something. Assigning to nirbhay_c for analysis.