[ https://issues.apache.org/jira/browse/HBASE-13259?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15156520#comment-15156520 ]
Anoop Sam John commented on HBASE-13259: ---------------------------------------- bq.buffers[i] = allocator.allocate(0, false); This is just a 0 length dummy buffer. Can be a HBB always. {code} if (directByteBuffer) { 77 buffer = fileChannel.map(java.nio.channels.FileChannel.MapMode.READ_WRITE, pos * size, 78 size); 79 } else { 80 buffer = fileChannel.map(java.nio.channels.FileChannel.MapMode.READ_WRITE, pos * size, 81 0); 82 } {code} On this engine, we will always call with 'directByteBuffer' true right? Can we throw Exception otherwise rather than creating 0 length BB? It is bit ugly to have sync() API in ByteBufferAllocator. {code} public void sync() throws IOException { 153 if (fileChannel != null) { 154 if(bufferArray != null) { 155 bufferArray.sync(); 156 } 157 } 158 } {code} Can be call force() on fileChannel? Depending on that, we may not need sync() in ByteBufferArray as well. // TODO: See if the SHARED mode can be created here You checked. I think it is not advisable. It depends on the total data size and the page hit/miss is going to decide the perf. > mmap() based BucketCache IOEngine > --------------------------------- > > Key: HBASE-13259 > URL: https://issues.apache.org/jira/browse/HBASE-13259 > Project: HBase > Issue Type: New Feature > Components: BlockCache > Affects Versions: 0.98.10 > Reporter: Zee Chen > Assignee: Zee Chen > Priority: Critical > Fix For: 2.0.0, 1.3.0 > > Attachments: HBASE-13259-v2.patch, HBASE-13259.patch, > HBASE-13259_v3.patch, ioread-1.svg, mmap-0.98-v1.patch, mmap-1.svg, > mmap-trunk-v1.patch > > > Of the existing BucketCache IOEngines, FileIOEngine uses pread() to copy data > from kernel space to user space. This is a good choice when the total working > set size is much bigger than the available RAM and the latency is dominated > by IO access. However, when the entire working set is small enough to fit in > the RAM, using mmap() (and subsequent memcpy()) to move data from kernel > space to user space is faster. I have run some short keyval gets tests and > the results indicate a reduction of 2%-7% of kernel CPU on my system, > depending on the load. On the gets, the latency histograms from mmap() are > identical to those from pread(), but peak throughput is close to 40% higher. > This patch modifies ByteByfferArray to allow it to specify a backing file. > Example for using this feature: set hbase.bucketcache.ioengine to > mmap:/dev/shm/bucketcache.0 in hbase-site.xml. > Attached perf measured CPU usage breakdown in flames graph. -- This message was sent by Atlassian JIRA (v6.3.4#6332)