[
https://issues.apache.org/jira/browse/KAFKA-3973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15411821#comment-15411821
]
Bill Bejeck edited comment on KAFKA-3973 at 8/8/16 3:34 PM:
------------------------------------------------------------
I used JMH to benchmark the performance of caching bytes vs object (tracking by
memory size using jamm) here are the results:
EDIT: New results from updated test
# Run complete. Total time: 00:02:41
Benchmark
Mode Cnt Score Error Units
MemoryBytesCacheBenchmark.testCacheByMemory thrpt 40 536694.504 ±
4177.019 ops/s
MemoryBytesCacheBenchmark.testCacheBySizeBytes thrpt 40 4713360.286 ±
60874.723 ops/s
Using JMH it still appears that serialization has the advantage.
The test used for benchmarking will be included in the PR for KAFKA-3989
(coming soon).
was (Author: bbejeck):
I used JMH to benchmark the performance of caching bytes vs object (tracking by
memory size using jamm) here are the results:
Result "testCacheBySizeBytes":
2157013.372 ±(99.9%) 198793.816 ops/s [Average]
(min, avg, max) = (687952.309, 2157013.372, 2485954.624), stdev = 353355.834
CI (99.9%): [1958219.556, 2355807.189] (assumes normal distribution)
# Run complete. Total time: 00:02:41
Benchmark
Mode Cnt Score Error Units
MemoryBytesCacheBenchmark.testCacheByMemory thrpt 40 290142.181 ±
3001.345 ops/s
MemoryBytesCacheBenchmark.testCacheBySizeBytes thrpt 40 2157013.372 ±
198793.816 ops/s
Using JMH it still appears that serialization has the advantage.
The test used for benchmarking will be included in the PR for KAFKA-3989
(coming soon).
> Investigate feasibility of caching bytes vs. records
> ----------------------------------------------------
>
> Key: KAFKA-3973
> URL: https://issues.apache.org/jira/browse/KAFKA-3973
> Project: Kafka
> Issue Type: Sub-task
> Components: streams
> Reporter: Eno Thereska
> Assignee: Bill Bejeck
> Fix For: 0.10.1.0
>
> Attachments: CachingPerformanceBenchmarks.java, MemoryLRUCache.java
>
>
> Currently the cache stores and accounts for records, not bytes or objects.
> This investigation would be around measuring any performance overheads that
> come from storing bytes or objects. As an outcome we should know whether 1)
> we should store bytes or 2) we should store objects.
> If we store objects, the cache still needs to know their size (so that it can
> know if the object fits in the allocated cache space, e.g., if the cache is
> 100MB and the object is 10MB, we'd have space for 10 such objects). The
> investigation needs to figure out how to find out the size of the object
> efficiently in Java.
> If we store bytes, then we are serialising an object into bytes before
> caching it, i.e., we take a serialisation cost. The investigation needs
> measure how bad this cost can be especially for the case when all objects fit
> in cache (and thus any extra serialisation cost would show).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
