[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[ https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16328646#comment-16328646 ] Pavel Kuznetsov edited comment on IGNITE-6217 at 1/18/18 9:26 AM: -- 1) Renamed 2) Cleaned up 3) To have more stable results. Since table size affects performance of query execution, we want `test` method to be executed with the same table size. [~vozerov] , correct me if I'm wrong 4) I don't mind, [~vozerov] , should we remove this? 5) -If I got it right, the most interesting results are native vs thin driver.- [~vozerov] -do we need v2 configurations ?- Yes we need this was (Author: pkouznet): 1) Renamed 2) Cleaned up 3) To have more stable results. Since table size affects performance of query execution, we want `test` method to be executed with the same table size. [~vozerov] , correct me if I'm wrong 4) I don't mind, [~vozerov] , should we remove this? 5) If I got it right, the most interesting results are native vs thin driver. [~vozerov] do we need v2 configurations ? > Add benchmark to compare JDBC drivers and native SQL execution > -- > > Key: IGNITE-6217 > URL: https://issues.apache.org/jira/browse/IGNITE-6217 > Project: Ignite > Issue Type: Task > Components: jdbc, sql, yardstick >Affects Versions: 2.1 >Reporter: Taras Ledkov >Assignee: Pavel Kuznetsov >Priority: Major > > We have to compare performance of the native SQL execution (via Ignite SQL > API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC > thin client. -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16328571#comment-16328571
]
Taras Ledkov edited comment on IGNITE-6217 at 1/17/18 10:06 AM:
[~pkouznet], my comments
1. {{AbstractJdbcBenchmark#findThinAddr}} please don't use
[abbreviations|https://cwiki.apache.org/confluence/display/IGNITE/Coding+Guidelines#CodingGuidelines-Abbreviations]
for methods name.
2. {{NativeSqlInsertDeleteBenchmark}} unused local variable.
3. Why we have only combined insert + delete benchmark is it impossible to
separate it?
4. I suggest to remove configuration for select ALL range for all SELECT
benchmarks. This operation is long and benchmark run will be failed by timeout.
We have to patch yardstick to benchmark long operations.
5. I cannot find new configurations to perform *insert-delete* and *update*
benchmarks for JDBC v2.
was (Author: tledkov-gridgain):
[~pkouznet], my comments
1. {{AbstractJdbcBenchmark#findThinAddr}} please don't use
[abbreviations|https://cwiki.apache.org/confluence/display/IGNITE/Coding+Guidelines#CodingGuidelines-Abbreviations]
for methods name.
2. {{NativeSqlInsertDeleteBenchmark}} unused local variable.
3. Why we have only combined insert + delete benchmark is it impossible to
separate it?
4. I suggest to remove configuration for select ALL range for all SELECT
benchmarks. This operation is long and benchmark run will be failed.
We have to patch yardstick to benchmark long operations.
5. I cannot find new configurations to perform *insert-delete* and *update*
benchmarks for JDBC v2.
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Pavel Kuznetsov
>Priority: Major
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/23/17 8:07 AM:
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
The query is executed with 1 thread.
|| Benchmark || Ops /sec (1 row) || Ops/sec (1000 rows) || Ops / sec (1
rows) || Ops / sec (1M rows) ||
| Nativa SQL | 9385.99 | 903.53 | 106.77 | 0.931 |
| JDBC v2 | .48 ({color:#d04437}-5%{color}) | 782.23
({color:#d04437}-13%{color}) | 105.27 ({color:#d04437}-1%{color}) | 0.919
({color:#d04437}-1%{color}) |
| JDBC thin | 4367.87 ({color:#d04437}-53%{color}) | 673.16
({color:#d04437}-25%{color}) | 94.90 ({color:#d04437}-11%{color}) | 0.886
({color:#d04437}-5%{color}) |
The parameters {{pageSize}} and {{fetchSize}} are default (1024).
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row) || Ops/sec (1000 rows) || Ops / sec (1
rows) || Ops / sec (1M rows) ||
| Nativa SQL | 9385.99 | 903.53 | 106.77 | 0.931 |
| JDBC v2 | .48 ({color:#d04437}-5%{color}) | 782.23
({color:#d04437}-13%{color}) | 105.27 ({color:#d04437}-1%{color}) | 0.919
({color:#d04437}-1%{color}) |
| JDBC thin | 4367.87 ({color:#d04437}-53%{color}) | 673.16
({color:#d04437}-25%{color}) | 94.90 ({color:#d04437}-11%{color}) | 0.886
({color:#d04437}-5%{color}) |
The parameters {{pageSize}} and {{fetchSize}} are default (1024).
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/17/17 12:32 PM:
-
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row) || Ops/sec (1000 rows) || Ops / sec (1
rows) || Ops / sec (1M rows) ||
| Nativa SQL | 9385.99 | 903.53 | 106.77 | 0.931 |
| JDBC v2 | .48 ({color:#d04437}-5%{color}) | 782.23
({color:#d04437}-13%{color}) | 105.27 ({color:#d04437}-1%{color}) | 0.919
({color:#d04437}-1%{color}) |
| JDBC thin | 4367.87 ({color:#d04437}-53%{color}) | 673.16
({color:#d04437}-25%{color}) | 94.90 ({color:#d04437}-11%{color}) | 0.886
({color:#d04437}-5%{color}) |
The parameters {{pageSize}} and {{fetchSize}} are default (1024).
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row) || Ops/sec (1000 rows) || Ops / sec (1
rows) || Ops / sec (1M rows) ||
| Nativa SQL | 9385.99 | 903.53 | 106.77 | 0.931 |
| JDBC v2 | .48 ({color:#d04437}-5%{color}) | 782.23
({color:#d04437}-13%{color}) | 105.27 ({color:#d04437}-1%{color}) | 0.919
({color:#d04437}-1%{color}) |
| JDBC thin | 4367.87 ({color:#d04437}-53%{color}) | 673.16
({color:#d04437}-5%{color}) | 94.90 ({color:#d04437}-11%{color}) | 0.886
({color:#d04437}-5%{color}) |
The parameters {{pageSize}} and {{fetchSize}} are default (1024).
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/17/17 10:06 AM:
-
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row) || Ops/sec (1000 rows) || Ops / sec (1
rows) || Ops / sec (1M rows) ||
| Nativa SQL | 9385.99 | 903.53 | 106.77 | 0.931 |
| JDBC v2 | .48 ({color:#d04437}-5%{color}) | 782.23
({color:#d04437}-13%{color}) | 105.27 ({color:#d04437}-1%{color}) | 0.919
({color:#d04437}-1%{color}) |
| JDBC thin | 4367.87 ({color:#d04437}-53%{color}) | 673.16
({color:#d04437}-5%{color}) | 94.90 ({color:#d04437}-11%{color}) | 0.886
({color:#d04437}-5%{color}) |
The parameters {{pageSize}} and {{fetchSize}} are default (1024).
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row) || Ops/sec (1000 rows) || Ops / sec (1
rows) || Ops / sec (1M rows) ||
| Nativa SQL | 9385.99 | 903.53 | 106.77 | 0.931 |
| JDBC v2 | .48 ({color:#d04437}-5%{color}) | 782.23
({color:#d04437}-13%{color}) | 105.27 ({color:#d04437}-1%{color}) | 0.919
({color:#d04437}-1%{color}) |
| JDBC thin | 4367.87 ({color:#d04437}-53%{color}) | 673.16
({color:#d04437}-5%{color}) | 94.90 ({color:#d04437}-11%{color}) | 0.886
({color:#d04437}-5%{color}) |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/17/17 10:03 AM:
-
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row) || Ops/sec (1000 rows) || Ops / sec (1
rows) || Ops / sec (1M rows) ||
| Nativa SQL | 9385.99 | 903.53 | 106.77 | 0.931 |
| JDBC v2 | .48 ({color:#d04437}-5%{color}) | 782.23
({color:#d04437}-13%{color}) | 105.27 ({color:#d04437}-1%{color}) | 0.919
({color:#d04437}-1%{color}) |
| JDBC thin | 4367.87 ({color:#d04437}-53%{color}) | 673.16
({color:#d04437}-5%{color}) | 94.90 ({color:#d04437}-11%{color}) | 0.886
({color:#d04437}-5%{color}) |
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row) || Ops/sec (1000 rows) || Ops / sec (1
rows) || Ops / sec (1M rows) ||
| Nativa SQL | 9385.99 | 903.53 | 106.77 | 0.931 |
| JDBC v2 | .48 | 782.23 | 105.27| 0.919 |
| JDBC thin | 4367.87 | 673.16 | 94.90 | 0.886 |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/17/17 9:55 AM:
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row) || Ops/sec (1000 rows) || Ops / sec (1
rows) || Ops / sec (1M rows) ||
| Nativa SQL | 9385.99 | 903.53 | 106.77 | 0.931 |
| JDBC v2 | .48 | 782.23 | 105.27| 0.919 |
| JDBC thin | 4367.87 | 673.16 | 94.90 | 0.886 |
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row) ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/16/17 2:43 PM:
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row) ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row in result) ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/16/17 2:42 PM:
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec (1 row in result) ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/16/17 9:46 AM:
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains unique 1M rows.
|| Benchmark || Ops /sec ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two column {{id long, val long}}
The table contains unique 1M rows.
|| Benchmark || Ops /sec ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/16/17 9:46 AM:
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains 1M unique rows.
|| Benchmark || Ops /sec ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two columns: {{id long, val long}}
The table contains unique 1M rows.
|| Benchmark || Ops /sec ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/16/17 9:46 AM:
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two column {{id long, val long}}
The table contains unique 1M rows.
|| Benchmark || Ops /sec ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two column {{id long, val long}}
The table contains unique 1M rows.
|| Benchmark || Ops /sec ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16205619#comment-16205619
]
Taras Ledkov edited comment on IGNITE-6217 at 10/16/17 9:29 AM:
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
SQL query:
{{SELECT val FROM test WHERE id = }}
where TEST table contains two column {{id long, val long}}
The table contains unique 1M rows.
|| Benchmark || Ops /sec ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
was (Author: tledkov-gridgain):
The other benchmark result.
Configurations:
# Native SQL: 1 server, 1 client. SQL queries are execute4d on client on
separate host
# JDBCv2: 2 1 server, queries are executed on JDBC v2 client on separate host.
# JDBC thin: 1 server, 1 client, JDBC thin client runs on separate host and
connects to the client node. SQL queries are executed on JDBC client on
separate4 host.
|| Benchmark || Ops /sec ||
| Nativa SQL | 9385.99 |
| JDBC v2 | .48 |
| JDBC thin | 4367.87 |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16148745#comment-16148745
]
Taras Ledkov edited comment on IGNITE-6217 at 10/10/17 9:51 AM:
{{benchmark-jdbc-sql.properties}} yardstick configuration is created.
Benchmark results (4 servers, 1 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 97983.90 | 4522.63 |
| *JDBC v2* | 76926.95 ({color:#d04437}-21%{color}) | 3706.86
({color:#d04437}-18%{color}) |
| *JDBC thin* | 38264.54 ({color:#d04437}-61%){color} | 2753.05
({color:#d04437}-25%{color}) |
Benchmark results (4 servers, 8 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 15841.80 | 1842.78 |
| *JDBC v2* | 14209.00 ({color:#d04437}-10%{color}) | 1247.67
({color:#d04437}-32%{color}) |
| *JDBC thin* | 5130.02 ({color:#d04437}-68%){color} | 332.39
({color:#d04437}-82%{color}) |
was (Author: tledkov-gridgain):
{{benchmark-jdbc-sql.properties}} yardstick configuration is created.
Benchmark results (4 servers, 1 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 97983.90 | 4522.63 |
| *JDBC v2* | 76926.95 ({color:#d04437}-21%{color}) | 3706.86
({color:#d04437}-18%{color}) |
| *JDBC thin* | 38264.54 ({color:#d04437}-61%){color} | 2753.05
({color:#d04437}-25%{color}) |
Benchmark results (4 servers, 8 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 15841.80 | 1842.78 |
| *JDBC v2* | 14209.00 ({color:#d04437}-10%{color}) | 1247.67
({color:#d04437}-32%{color}) |
| *JDBC thin* | 5130.02 ({color:#d04437}-68%){color} | 332.39
({color:#d04437}-82%{color}) |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16148745#comment-16148745
]
Taras Ledkov edited comment on IGNITE-6217 at 10/10/17 9:51 AM:
{{benchmark-jdbc-sql.properties}} yardstick configuration is created.
Benchmark results (4 servers, 1 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 97983.90 | 4522.63 |
| *JDBC v2* | 76926.95 ({color:#d04437}-21%{color}) | 3706.86
({color:#d04437}-18%{color}) |
| *JDBC thin* | 38264.54 ({color:#d04437}-61%){color} | 2753.05
({color:#d04437}-25%{color}) |
Benchmark results (4 servers, 8 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 15841.80 | 1842.78 |
| *JDBC v2* | 14209.00 ({color:#d04437}-10%{color}) | 1247.67
({color:#d04437}-32%{color}) |
| *JDBC thin* | 5130.02 ({color:#d04437}-68%){color} | 332.39
({color:#d04437}-82%{color}) |
was (Author: tledkov-gridgain):
{{benchmark-jdbc-sql.properties}} yardstick configuration is created.
Benchmark results (4 servers, 1 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 97983.90 | 4522.63 |
| *JDBC v2* | 76926.95 ({color:#d04437}-21%{color}) | 3706.86
({color:#d04437}-18%{color}) |
| *JDBC thin* | 38264.54 ({color:#d04437}-61%){color} | 2753.05
({color:#d04437}-25%{color}) |
Benchmark results (4 servers, 8 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 15841.80 | 1842.78 |
| *JDBC v2* | 14209.00 ({color:#d04437}-10%{color}) | 1247.67
({color:#d04437}-32%{color}) |
| *JDBC thin* | 5130.02 ({color:#d04437}-68%){color} | 332.39
({color:#d04437}-82%{color}) |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16148745#comment-16148745
]
Taras Ledkov edited comment on IGNITE-6217 at 10/10/17 9:50 AM:
{{benchmark-jdbc-sql.properties}} yardstick configuration is created.
Benchmark results (4 servers, 1 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 97983.90 | 4522.63 |
| *JDBC v2* | 76926.95 ({color:#d04437}-21%{color}) | 3706.86
({color:#d04437}-18%{color}) |
| *JDBC thin* | 38264.54 ({color:#d04437}-61%){color} | 2753.05
({color:#d04437}-25%{color}) |
Benchmark results (4 servers, 8 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 15841.80 | 1842.78 |
| *JDBC v2* | 14209.00 ({color:#d04437}-10%{color}) | 1247.67
({color:#d04437}-32%{color}) |
| *JDBC thin* | 5130.02 ({color:#d04437}-68%){color} | 332.39
({color:#d04437}-82%{color}) |
was (Author: tledkov-gridgain):
{{benchmark-jdbc-sql.properties}} yardstick configuration is created.
Benchmark results (4 servers, 1 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 97983.90 | 4522.63 |
| *JDBC v2* | 76926.95 ({color:#d04437}-21%{color}) | 3706.86
({color:#d04437}-18%{color}) |
| *JDBC thin* | 38264.54 ({color:#d04437}-61%){color} | 2753.05
({color:#d04437}-25%{color}) |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16148745#comment-16148745
]
Taras Ledkov edited comment on IGNITE-6217 at 10/6/17 1:34 PM:
---
{{benchmark-jdbc-sql.properties}} yardstick configuration is created.
Benchmark results (4 servers, 1 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 97983.90 | 4522.63 |
| *JDBC v2* | 76926.95 ({color:#d04437}-21%{color}) | 3706.86
({color:#d04437}-18%{color}) |
| *JDBC thin* | 38264.54 ({color:#d04437}-61%){color} | 2753.05
({color:#d04437}-25%{color}) |
was (Author: tledkov-gridgain):
{{benchmark-jdbc-sql.properties}} yardstick configuration is created.
Benchmark results (4 servers, 1 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 97983.90 | 4522.63 |
| *JDBC v2* | 76926.95 ({color:#d04437}-21%{color}) | 3706.86
({color:#d04437}-18%{color}) |
| *JDBC thin* | 38264.54 ({color:#d04437}-61%){color} | 2753.05
({color:#d04437}-25%{color}) |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Comment Edited] (IGNITE-6217) Add benchmark to compare JDBC drivers and native SQL execution
[
https://issues.apache.org/jira/browse/IGNITE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16148745#comment-16148745
]
Taras Ledkov edited comment on IGNITE-6217 at 10/6/17 1:31 PM:
---
{{benchmark-jdbc-sql.properties}} yardstick configuration is created.
Benchmark results (4 servers, 1 client):
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 97983.90 | 4522.63 |
| *JDBC v2* | 76926.95 ({color:#d04437}-21%{color}) | 3706.86
({color:#d04437}-18%{color}) |
| *JDBC thin* | 38264.54 ({color:#d04437}-61%){color} | 2753.05
({color:#d04437}-25%{color}) |
was (Author: tledkov-gridgain):
{{benchmark-jdbc-sql.properties}} yardstick configuration is created.
Local benchmark results:
|| Benchmark || Ops/sec (1 row in results) || Ops/sec (1000 rows in results) ||
| *NativeSQL* | 8,413.40 | 266.85 |
| *JDBC v2* | 6,920.51 | 186.76 |
| *JDBC thin* | 2,672.72 | 148.20 |
> Add benchmark to compare JDBC drivers and native SQL execution
> --
>
> Key: IGNITE-6217
> URL: https://issues.apache.org/jira/browse/IGNITE-6217
> Project: Ignite
> Issue Type: Task
> Components: jdbc, sql, yardstick
>Affects Versions: 2.1
>Reporter: Taras Ledkov
>Assignee: Taras Ledkov
> Fix For: 2.4
>
>
> We have to compare performance of the native SQL execution (via Ignite SQL
> API), JDBC v2 driver, that uses Ignite client to connect to grid and JDBC
> thin client.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
