[
https://issues.apache.org/jira/browse/CASSANDRA-21536?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dmitry Konstantinov reassigned CASSANDRA-21536:
-----------------------------------------------
Assignee: koo
> Profile pollution in AbstractType.writeValue makes serialization slow for all
> column types
> ------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-21536
> URL: https://issues.apache.org/jira/browse/CASSANDRA-21536
> Project: Apache Cassandra
> Issue Type: Improvement
> Reporter: koo
> Assignee: koo
> Priority: Normal
> Fix For: 6.x, 7.x
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> *Description:*
> AbstractType.writeValue() is one shared method. All column types use it.
> Inside writeValue(), it calls valueLengthIfFixed(). This is a virtual call.
> Many types override this method (Int32Type, LongType, UTF8Type, ...).
> In a real cluster, many column types pass through writeValue(). So the
> profile always sees lots types.
> Because of this, the JIT cannot inline the call. It stays as a vtable call.
> Also, the compiled body of writeValue() becomes big, so the JIT refuses to
> inline writeValue() itself ("already compiled into a big method").
> We also see the same itable/vtable stubs in async-profiler output from a real
> cluster, running with default production options and a normal workload.
> *How to solve:*
> Add a final int field to AbstractType. Set it in the constructor.
> writeValue() reads this field instead of calling valueLengthIfFixed().
> A field read needs no type profile. So profile pollution has no effect on it.
> The valueLengthIfFixed() method is not changed. Only the write path uses the
> field.
> *Result:*
> Production is always the polluted state, so this is the real-world comparison
> || || before || after ||
> | M ops/s | 185.0 | 210.4 |
> | improvement | | +13.7% |
> | valueLengthIfFixed frames in profiler | 5.9% | 0% |
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
