[ https://issues.apache.org/jira/browse/HBASE-20188?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16424604#comment-16424604 ]
Eshcar Hillel commented on HBASE-20188: --------------------------------------- Attached are the results of evaluations over *SSD* machines [^HBase 2.0 performance evaluation - Basic vs None_ system settings.pdf] , and the script to run them [^HBASE-20188.sh] (which is based on the script by Stack). The setting is also similar: 1 master, 1RS with 8GB heap, 1 ycsb client, underlying HDFS set to 3-way replication. Comparing Basic with default configuration vs None under different system settings: cms/mslab vs cms/no-mslabs vs g1gc/no-maslab Summary of results: 1) None outperforms Basic in a uniform distribution of insert-only operations that includes multiple split events 2) Basic outperforms None in a mixed workload with zipfian distribution 3) None is slightly better than Basic in read-only zipfian workload 4) not using mslab improves performance in zipfian distribution workloads and has a negative effect with insert-only uniform workload 5) g1gc performs worse in all cases; this could be due to lack of tuning It is important to note that each configuration was tested once, each of these runs can be an outlier - a good or bad outlier Next we will come up with a workload which demonstrates the advantage of in-memory compaction as well as continue with benchmarks to determine optimal default values for in-memory compaction, namely portion of active segment, length of pipeline, etc. > [TESTING] Performance > --------------------- > > Key: HBASE-20188 > URL: https://issues.apache.org/jira/browse/HBASE-20188 > Project: HBase > Issue Type: Umbrella > Components: Performance > Reporter: stack > Assignee: stack > Priority: Blocker > Fix For: 2.0.0 > > Attachments: CAM-CONFIG-V01.patch, HBASE-20188.sh, HBase 2.0 > performance evaluation - Basic vs None_ system settings.pdf, > ITBLL2.5B_1.2.7vs2.0.0_cpu.png, ITBLL2.5B_1.2.7vs2.0.0_gctime.png, > ITBLL2.5B_1.2.7vs2.0.0_iops.png, ITBLL2.5B_1.2.7vs2.0.0_load.png, > ITBLL2.5B_1.2.7vs2.0.0_memheap.png, ITBLL2.5B_1.2.7vs2.0.0_memstore.png, > ITBLL2.5B_1.2.7vs2.0.0_ops.png, > ITBLL2.5B_1.2.7vs2.0.0_ops_NOT_summing_regions.png, YCSB_CPU.png, > YCSB_GC_TIME.png, YCSB_IN_MEMORY_COMPACTION=NONE.ops.png, YCSB_MEMSTORE.png, > YCSB_OPs.png, YCSB_in-memory-compaction=NONE.ops.png, YCSB_load.png, > flamegraph-1072.1.svg, flamegraph-1072.2.svg, tree.txt > > > How does 2.0.0 compare to old versions? Is it faster, slower? There is rumor > that it is much slower, that the problem is the asyncwal writing. Does > in-memory compaction slow us down or speed us up? What happens when you > enable offheaping? > Keep notes here in this umbrella issue. Need to be able to say something > about perf when 2.0.0 ships. -- This message was sent by Atlassian JIRA (v7.6.3#76005)