[jira] [Commented] (IGNITE-20617) SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
[
https://issues.apache.org/jira/browse/IGNITE-20617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17782326#comment-17782326
]
Ivan Artiukhov commented on IGNITE-20617:
-
[~korlov]
AI3, rev. a9ed167a5045ba166a6b3c84f8fb9df486579b14 (Nov 2 2023)
I've run the same test you run, i.e. with 2 server nodes and with 2 IP
addresses passed to thin client.
{code:java}
Command line: -db site.ycsb.db.ignite3.IgniteSqlClient -t -P
/opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
operationcount=1 -p recordcount=1 -p warmupops=2000 -p
dataintegrity=true -p measurementtype=timeseries -p status.interval=1 -p
hosts=192.168.1.122,192.168.1.34 -p recordcount=20 -p operationcount=20
-s
YCSB Client 2023.7{code}
Unfortunately I see far different performance than on your graph.
!2023-11-02-15-41-10_run.png!
The benchmark log: [^2023-11-02-15-41-10_run.txt]
Server nodes logs, JFR, iostat: [^2023-11-02-15-41-10_logs.zip]
> SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
>
>
> Key: IGNITE-20617
> URL: https://issues.apache.org/jira/browse/IGNITE-20617
> Project: Ignite
> Issue Type: Bug
> Components: sql
>Reporter: Ivan Artiukhov
>Assignee: Konstantin Orlov
>Priority: Major
> Labels: ignite-3, performance
> Attachments: 2023-11-02-15-41-10_logs.zip,
> 2023-11-02-15-41-10_run.png, 2023-11-02-15-41-10_run.txt,
> 6b6f0c-select-2srv.png, 6b6f0c-select-logs.zip.zip, echarts.png,
> ignite-config.json, jdbc-1node-select.png, jdbc-2nodes-select.png,
> sql-1node-select.png, sql-2nodes-select.png, sql-logs-2-server-nodes.zip
>
>
> Ignite 3, rev. 7d188ac7ae068bd69ff0e6e6cfe5a32ac5749d17
> Benchmark:
> [https://github.com/gridgain/YCSB/blob/ycsb-2023.3/ignite3/src/main/java/site/ycsb/db/ignite3/IgniteSqlClient.java]
>
> The benchmark establishes an SQL {{Session}} and perform {{SELECTs}} via
> {{{}preparedStatement{}}}.
>
> Steps:
> * Run an Ignite cluster of 2 nodes with the attached config
> [^ignite-config.json] .
> ** *fsync = false*
> * Run the SQL YCSB benchmark in preload mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -load -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> recordcount=1 -p dataintegrity=true -p measurementtype=timeseries -p
> hosts=192.168.1.60}}
> * Run the SQL YCSB benchmark in 100% read mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -t -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> operationcount=1 -p recordcount=1 -p dataintegrity=true -p
> measurementtype=timeseries -p hosts=192.168.1.60 -s}}
> * Observe the following average throughput on reads:
> !sql-2nodes-select.png!
> Server node's logs: [^sql-logs-2-server-nodes.zip]
> Repeat the test with only 1 server node and observe {*}~20x better throughput
> on reads{*}:
> !sql-1node-select.png!
>
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
[jira] [Commented] (IGNITE-20617) SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
[
https://issues.apache.org/jira/browse/IGNITE-20617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17782180#comment-17782180
]
Ivan Artiukhov commented on IGNITE-20617:
-
[~korlov] I see that you run the benchmarks with 2 server node IPs passed to
the client.
My results are also for 2 nodes, but I passed 1 server node IP address.
I'll check one more time.
> SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
>
>
> Key: IGNITE-20617
> URL: https://issues.apache.org/jira/browse/IGNITE-20617
> Project: Ignite
> Issue Type: Bug
> Components: sql
>Reporter: Ivan Artiukhov
>Assignee: Konstantin Orlov
>Priority: Major
> Labels: ignite-3, performance
> Attachments: 6b6f0c-select-2srv.png, 6b6f0c-select-logs.zip.zip,
> echarts.png, ignite-config.json, jdbc-1node-select.png,
> jdbc-2nodes-select.png, sql-1node-select.png, sql-2nodes-select.png,
> sql-logs-2-server-nodes.zip
>
>
> Ignite 3, rev. 7d188ac7ae068bd69ff0e6e6cfe5a32ac5749d17
> Benchmark:
> [https://github.com/gridgain/YCSB/blob/ycsb-2023.3/ignite3/src/main/java/site/ycsb/db/ignite3/IgniteSqlClient.java]
>
> The benchmark establishes an SQL {{Session}} and perform {{SELECTs}} via
> {{{}preparedStatement{}}}.
>
> Steps:
> * Run an Ignite cluster of 2 nodes with the attached config
> [^ignite-config.json] .
> ** *fsync = false*
> * Run the SQL YCSB benchmark in preload mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -load -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> recordcount=1 -p dataintegrity=true -p measurementtype=timeseries -p
> hosts=192.168.1.60}}
> * Run the SQL YCSB benchmark in 100% read mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -t -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> operationcount=1 -p recordcount=1 -p dataintegrity=true -p
> measurementtype=timeseries -p hosts=192.168.1.60 -s}}
> * Observe the following average throughput on reads:
> !sql-2nodes-select.png!
> Server node's logs: [^sql-logs-2-server-nodes.zip]
> Repeat the test with only 1 server node and observe {*}~20x better throughput
> on reads{*}:
> !sql-1node-select.png!
>
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
[jira] [Commented] (IGNITE-20617) SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
[
https://issues.apache.org/jira/browse/IGNITE-20617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17781390#comment-17781390
]
Konstantin Orlov commented on IGNITE-20617:
---
[~Artukhov], sorry for bothering you, but could you double check please?
I've run this benchmark on 2 server node with following run configuration
{code:java}
Command line: -db site.ycsb.db.ignite3.IgniteSqlClient -t -P
/opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
operationcount=1 -p recordcount=1 -p warmupops=2000 -p
dataintegrity=true -p measurementtype=timeseries -p status.interval=1 -p
hosts=192.168.1.41,192.168.1.6 -p operationcount=10 -p recordcount=10 -s
YCSB Client 2023.7
{code}
and results looks normal to me:
!echarts.png!
Although I run benchmark on fresh main
(753dcde0095e0061efddc15a61c0ae2dc8c80239), I've also checked history between
two revisions, and have found no commits that potentially may address this issue
> SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
>
>
> Key: IGNITE-20617
> URL: https://issues.apache.org/jira/browse/IGNITE-20617
> Project: Ignite
> Issue Type: Bug
> Components: sql
>Reporter: Ivan Artiukhov
>Assignee: Konstantin Orlov
>Priority: Major
> Labels: ignite-3, performance
> Attachments: 6b6f0c-select-2srv.png, 6b6f0c-select-logs.zip.zip,
> echarts.png, ignite-config.json, jdbc-1node-select.png,
> jdbc-2nodes-select.png, sql-1node-select.png, sql-2nodes-select.png,
> sql-logs-2-server-nodes.zip
>
>
> Ignite 3, rev. 7d188ac7ae068bd69ff0e6e6cfe5a32ac5749d17
> Benchmark:
> [https://github.com/gridgain/YCSB/blob/ycsb-2023.3/ignite3/src/main/java/site/ycsb/db/ignite3/IgniteSqlClient.java]
>
> The benchmark establishes an SQL {{Session}} and perform {{SELECTs}} via
> {{{}preparedStatement{}}}.
>
> Steps:
> * Run an Ignite cluster of 2 nodes with the attached config
> [^ignite-config.json] .
> ** *fsync = false*
> * Run the SQL YCSB benchmark in preload mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -load -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> recordcount=1 -p dataintegrity=true -p measurementtype=timeseries -p
> hosts=192.168.1.60}}
> * Run the SQL YCSB benchmark in 100% read mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -t -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> operationcount=1 -p recordcount=1 -p dataintegrity=true -p
> measurementtype=timeseries -p hosts=192.168.1.60 -s}}
> * Observe the following average throughput on reads:
> !sql-2nodes-select.png!
> Server node's logs: [^sql-logs-2-server-nodes.zip]
> Repeat the test with only 1 server node and observe {*}~20x better throughput
> on reads{*}:
> !sql-1node-select.png!
>
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
[jira] [Commented] (IGNITE-20617) SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
[
https://issues.apache.org/jira/browse/IGNITE-20617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17780404#comment-17780404
]
Ivan Artiukhov commented on IGNITE-20617:
-
AI3, rev. 6b6f0cb6109d7906a980a31522a1b9c36f88e43a (Fri Oct 27)
The same benchmark still shows poor SELECTs performance on 2 server nodes.
!6b6f0c-select-2srv.png!
Nodes logs, GC logs: [^6b6f0c-select-logs.zip.zip]
> SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
>
>
> Key: IGNITE-20617
> URL: https://issues.apache.org/jira/browse/IGNITE-20617
> Project: Ignite
> Issue Type: Bug
> Components: sql
>Reporter: Ivan Artiukhov
>Assignee: Konstantin Orlov
>Priority: Major
> Labels: ignite-3, performance
> Attachments: 6b6f0c-select-2srv.png, 6b6f0c-select-logs.zip.zip,
> ignite-config.json, jdbc-1node-select.png, jdbc-2nodes-select.png,
> sql-1node-select.png, sql-2nodes-select.png, sql-logs-2-server-nodes.zip
>
>
> Ignite 3, rev. 7d188ac7ae068bd69ff0e6e6cfe5a32ac5749d17
> Benchmark:
> [https://github.com/gridgain/YCSB/blob/ycsb-2023.3/ignite3/src/main/java/site/ycsb/db/ignite3/IgniteSqlClient.java]
>
> The benchmark establishes an SQL {{Session}} and perform {{SELECTs}} via
> {{{}preparedStatement{}}}.
>
> Steps:
> * Run an Ignite cluster of 2 nodes with the attached config
> [^ignite-config.json] .
> ** *fsync = false*
> * Run the SQL YCSB benchmark in preload mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -load -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> recordcount=1 -p dataintegrity=true -p measurementtype=timeseries -p
> hosts=192.168.1.60}}
> * Run the SQL YCSB benchmark in 100% read mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -t -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> operationcount=1 -p recordcount=1 -p dataintegrity=true -p
> measurementtype=timeseries -p hosts=192.168.1.60 -s}}
> * Observe the following average throughput on reads:
> !sql-2nodes-select.png!
> Server node's logs: [^sql-logs-2-server-nodes.zip]
> Repeat the test with only 1 server node and observe {*}~20x better throughput
> on reads{*}:
> !sql-1node-select.png!
>
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
[jira] [Commented] (IGNITE-20617) SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
[
https://issues.apache.org/jira/browse/IGNITE-20617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17776116#comment-17776116
]
Konstantin Orlov commented on IGNITE-20617:
---
Temporal fix should be introduced in IGNITE-20668
> SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
>
>
> Key: IGNITE-20617
> URL: https://issues.apache.org/jira/browse/IGNITE-20617
> Project: Ignite
> Issue Type: Bug
> Components: sql
>Reporter: Ivan Artiukhov
>Priority: Major
> Labels: ignite-3, performance
> Attachments: ignite-config.json, jdbc-1node-select.png,
> jdbc-2nodes-select.png, sql-1node-select.png, sql-2nodes-select.png,
> sql-logs-2-server-nodes.zip
>
>
> Ignite 3, rev. 7d188ac7ae068bd69ff0e6e6cfe5a32ac5749d17
> Benchmark:
> [https://github.com/gridgain/YCSB/blob/ycsb-2023.3/ignite3/src/main/java/site/ycsb/db/ignite3/IgniteSqlClient.java]
>
> The benchmark establishes an SQL {{Session}} and perform {{SELECTs}} via
> {{{}preparedStatement{}}}.
>
> Steps:
> * Run an Ignite cluster of 2 nodes with the attached config
> [^ignite-config.json] .
> ** *fsync = false*
> * Run the SQL YCSB benchmark in preload mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -load -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> recordcount=1 -p dataintegrity=true -p measurementtype=timeseries -p
> hosts=192.168.1.60}}
> * Run the SQL YCSB benchmark in 100% read mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -t -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> operationcount=1 -p recordcount=1 -p dataintegrity=true -p
> measurementtype=timeseries -p hosts=192.168.1.60 -s}}
> * Observe the following average throughput on reads:
> !sql-2nodes-select.png!
> Server node's logs: [^sql-logs-2-server-nodes.zip]
> Repeat the test with only 1 server node and observe {*}~20x better throughput
> on reads{*}:
> !sql-1node-select.png!
>
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
[jira] [Commented] (IGNITE-20617) SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
[
https://issues.apache.org/jira/browse/IGNITE-20617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17776108#comment-17776108
]
Konstantin Orlov commented on IGNITE-20617:
---
The reason of such drastic degradation is code in
{{TxManagerImpl#currentReadTimestamp()}} that was commented out under {{TODO:
IGNITE-20378 Fix it}}. Because of that, RO transaction doesn't start in the
past as it should, which results in a waiting of a safe time on every read.
> SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
>
>
> Key: IGNITE-20617
> URL: https://issues.apache.org/jira/browse/IGNITE-20617
> Project: Ignite
> Issue Type: Bug
> Components: sql
>Reporter: Ivan Artiukhov
>Priority: Major
> Labels: ignite-3, performance
> Attachments: ignite-config.json, jdbc-1node-select.png,
> jdbc-2nodes-select.png, sql-1node-select.png, sql-2nodes-select.png,
> sql-logs-2-server-nodes.zip
>
>
> Ignite 3, rev. 7d188ac7ae068bd69ff0e6e6cfe5a32ac5749d17
> Benchmark:
> [https://github.com/gridgain/YCSB/blob/ycsb-2023.3/ignite3/src/main/java/site/ycsb/db/ignite3/IgniteSqlClient.java]
>
> The benchmark establishes an SQL {{Session}} and perform {{SELECTs}} via
> {{{}preparedStatement{}}}.
>
> Steps:
> * Run an Ignite cluster of 2 nodes with the attached config
> [^ignite-config.json] .
> ** *fsync = false*
> * Run the SQL YCSB benchmark in preload mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -load -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> recordcount=1 -p dataintegrity=true -p measurementtype=timeseries -p
> hosts=192.168.1.60}}
> * Run the SQL YCSB benchmark in 100% read mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -t -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> operationcount=1 -p recordcount=1 -p dataintegrity=true -p
> measurementtype=timeseries -p hosts=192.168.1.60 -s}}
> * Observe the following average throughput on reads:
> !sql-2nodes-select.png!
> Server node's logs: [^sql-logs-2-server-nodes.zip]
> Repeat the test with only 1 server node and observe {*}~20x better throughput
> on reads{*}:
> !sql-1node-select.png!
>
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
[jira] [Commented] (IGNITE-20617) SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
[
https://issues.apache.org/jira/browse/IGNITE-20617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17773929#comment-17773929
]
Ivan Artiukhov commented on IGNITE-20617:
-
The similar drop is observed on the following JDBC benchmark:
[https://github.com/gridgain/YCSB/blob/ycsb-2023.3/ignite3/src/main/java/site/ycsb/db/ignite3/IgniteJdbcClient.java]
100% read on 2 nodes cluster:
!jdbc-2nodes-select.png!
100% read on 1 node cluster:
!jdbc-1node-select.png!
> SQL: ~20x performance degradation in SELECTS (2 nodes VS 1 node)
>
>
> Key: IGNITE-20617
> URL: https://issues.apache.org/jira/browse/IGNITE-20617
> Project: Ignite
> Issue Type: Bug
> Components: sql
>Reporter: Ivan Artiukhov
>Priority: Major
> Labels: ignite-3, performance
> Attachments: ignite-config.json, jdbc-1node-select.png,
> jdbc-2nodes-select.png, sql-1node-select.png, sql-2nodes-select.png
>
>
> Ignite 3, rev. 7d188ac7ae068bd69ff0e6e6cfe5a32ac5749d17
> Benchmark:
> [https://github.com/gridgain/YCSB/blob/ycsb-2023.3/ignite3/src/main/java/site/ycsb/db/ignite3/IgniteSqlClient.java]
>
> The benchmark establishes an SQL {{Session}} and perform {{SELECTs}} via
> {{{}preparedStatement{}}}.
>
> Steps:
> * Run an Ignite cluster of 2 nodes with the attached config
> [^ignite-config.json]
> * Run the SQL YCSB benchmark in preload mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -load -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> recordcount=1 -p dataintegrity=true -p measurementtype=timeseries -p
> hosts=192.168.1.60}}
> * Run the SQL YCSB benchmark in 100% read mode:
> ** {{-db site.ycsb.db.ignite3.IgniteSqlClient -t -P
> /opt/pubagent/poc/config/ycsb/workloads/workloadc -threads 1 -p
> operationcount=1 -p recordcount=1 -p dataintegrity=true -p
> measurementtype=timeseries -p hosts=192.168.1.60 -s}}
> * Observe the following average throughput on reads:
> !sql-2nodes-select.png!
> Repeat the test with only 1 server node and observe {*}~20x better throughput
> on reads{*}:
> !sql-1node-select.png!
>
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
