[
https://issues.apache.org/jira/browse/HBASE-20751?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16523274#comment-16523274
]
Zheng Hu edited comment on HBASE-20751 at 6/26/18 6:38 AM:
-----------------------------------------------------------
Uploaded the testing result, we have 3 cases:
1. master branch with the async replication enabled.
2. HBASE-19064 branch with async replication enabled.
3. HBASE-19064 branch with sync replication enabled.
For each case, I've created a table with 100 regions, and used YCSB(bin/ycsb
run hbase10 -P workload -s -threads 120) with 120 threads to write the source
hbase cluster.
{code}
hbase> n_splits = 100
hbase> create 'ycsb-test', {NAME=>'C', REPLICATION_SCOPE=>'1'}, {SPLITS =>
(1..n_splits).map {|i| "user#{1000+i*(9999-1000)/n_splits}"}}
{code}
The ycsb workload:
{code}
table=ycsb-test
columnfamily=C
recordcount=100000000
operationcount=100000000
workload=com.yahoo.ycsb.workloads.CoreWorkload
fieldlength=100
fieldcount=1
autoflush=false
clientbuffering=false # Ensure that using PUT, rather than BufferMutator.
BTW, use the default duraibility=SYNC_WAL.
readallfields=true
writeallfields=true
readproportion=0
updateproportion=0
scanproportion=0
insertproportion=1.0
requestdistribution=zipfian
{code}
was (Author: openinx):
Uploaded the testing result, we have 3 cases:
1. master branch with the async replication enabled.
2. HBASE-19064 branch with async replication enabled.
3. HBASE-19064 branch with sync replication enabled.
For each case, I've created a table with 100 regions, and used YCSBbin/ycsb
run hbase10 -P workload -s -threads 120) with 120 threads to write the source
hbase cluster.
{code}
hbase> n_splits = 100
hbase> create 'ycsb-test', {NAME=>'C', REPLICATION_SCOPE=>'1'}, {SPLITS =>
(1..n_splits).map {|i| "user#{1000+i*(9999-1000)/n_splits}"}}
{code}
The ycsb workload:
{code}
table=ycsb-test
columnfamily=C
recordcount=100000000
operationcount=100000000
workload=com.yahoo.ycsb.workloads.CoreWorkload
fieldlength=100
fieldcount=1
autoflush=false
clientbuffering=false # Ensure that using PUT, rather than BufferMutator.
BTW, use the default duraibility=SYNC_WAL.
readallfields=true
writeallfields=true
readproportion=0
updateproportion=0
scanproportion=0
insertproportion=1.0
requestdistribution=zipfian
{code}
> Performance comparison of synchronous replication branch and master branch
> ---------------------------------------------------------------------------
>
> Key: HBASE-20751
> URL: https://issues.apache.org/jira/browse/HBASE-20751
> Project: HBase
> Issue Type: Sub-task
> Reporter: Zheng Hu
> Assignee: Zheng Hu
> Priority: Major
> Attachments: HBASE-19064-branch-async-replication.png,
> HBASE-19064-branch-sync-replication.png,
> master-branch-async-replication-put100000000.png, result.png, ycsb-log.tar.gz
>
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
