terrymanu commented on issue #34771:
URL:
https://github.com/apache/shardingsphere/issues/34771#issuecomment-3506426715
After detailed analysis, this issue is not a bug in ShardingSphere, but is
caused by incorrect configuration.
Root Cause Analysis
The fundamental problem is that the read-write splitting rule names do not
match the data source names after sharding routing:
1. Your Initial Configuration:
ReadwriteSplittingDataSourceRuleConfiguration readWriteSource0 = new
ReadwriteSplittingDataSourceRuleConfiguration(
"write_read_split0", "master0", Collections.singletonList("save0"),
TransactionalReadQueryStrategy.PRIMARY, "roundRobin");
2. Routing Execution Flow:
- Sharding routing executes first, routing queries to actual data
sources (e.g., master1)
- Read-write splitting routing executes second, needing to find
corresponding read-write splitting rules based on data source names
- Your sharding routing results in data source names: master0, master1,
master2
- But your read-write splitting rule names are: write_read_split0,
write_read_split1, write_read_split2
3. Matching Failure:
// This matching logic fails because "master1" != "write_read_split1"
if (dataSourceRule.isPresent() &&
((ReadwriteSplittingDataSourceRule)dataSourceRule.get()).getName().equalsIgnoreCase(each.getDataSourceMapper().getActualName()))
{
// This branch is not entered
}
Correct Configuration
Your solution is correct. The read-write splitting rule names must match
the data source names after sharding routing:
ReadwriteSplittingDataSourceRuleConfiguration readWriteSource0 = new
ReadwriteSplittingDataSourceRuleConfiguration(
"master0", "master0", Collections.singletonList("save0"),
TransactionalReadQueryStrategy.PRIMARY, "roundRobin");
ReadwriteSplittingDataSourceRuleConfiguration readWriteSource1 = new
ReadwriteSplittingDataSourceRuleConfiguration(
"master1", "master1", Collections.singletonList("save1"),
TransactionalReadQueryStrategy.PRIMARY, "roundRobin");
ReadwriteSplittingDataSourceRuleConfiguration readWriteSource2 = new
ReadwriteSplittingDataSourceRuleConfiguration(
"master2", "master2", Collections.singletonList("save2"),
TransactionalReadQueryStrategy.PRIMARY, "roundRobin");
Parameter Description
- First Parameter: Read-write splitting rule name, must match the actual
data source name after sharding routing
- Second Parameter: Write data source name
- Third Parameter: Read data source list
- Fourth Parameter: Read strategy within transactions
- Fifth Parameter: Load balancing algorithm
Conclusion
This issue aligns with ShardingSphere's design principles. When using both
sharding and read-write splitting features simultaneously, the read-write
splitting rules need to recognize the results of sharding routing,
therefore the rule names must match.
Your final solution is completely correct. It is recommended to continue
using this configuration approach.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
