[
https://issues.apache.org/jira/browse/TAJO-623?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
DaeMyung Kang updated TAJO-623:
-------------------------------
Description:
Currently, to chnage network buffer size or tcp_nodelay, or keepalive options,
we should change source code.
```c
<property>
<name>tajo.network.receivebuffersize</name>
<value>10485760</value>
</property>
```
NETWORK_REUSEADDRESS("tajo.network.reuseaddress", true)
NETWORK_TCP_NODELAY("tajo.network.tcpnodelay", true)
NETWORK_KEEP_ALIVE("tajo.network.keepalive", true)
NETWORK_CONNECT_TIMEOUT_MILLIS("tajo.network.connecttimeoutmillis", 10000)
NETWORK_CONNECT_RESPONSE_TIMEOUT_MILLIS("tajo.network.connectresponsetimeoutmillis",
10000)
NETWORK_RECEIVE_BUFFERSIZE("tajo.network.receivebuffersize", 1048576 * 10)
was:
Currently, to chnage network buffer size or tcp_nodelay, or keepalive options,
we should change source code.
> Extract some network setting value to conf
> ------------------------------------------
>
> Key: TAJO-623
> URL: https://issues.apache.org/jira/browse/TAJO-623
> Project: Tajo
> Issue Type: Improvement
> Components: conf and scripts
> Reporter: DaeMyung Kang
> Assignee: DaeMyung Kang
> Priority: Minor
> Attachments: TAJO-623.patch
>
>
> Currently, to chnage network buffer size or tcp_nodelay, or keepalive
> options, we should change source code.
> ```c
> <property>
> <name>tajo.network.receivebuffersize</name>
> <value>10485760</value>
> </property>
> ```
> NETWORK_REUSEADDRESS("tajo.network.reuseaddress", true)
> NETWORK_TCP_NODELAY("tajo.network.tcpnodelay", true)
> NETWORK_KEEP_ALIVE("tajo.network.keepalive", true)
> NETWORK_CONNECT_TIMEOUT_MILLIS("tajo.network.connecttimeoutmillis", 10000)
> NETWORK_CONNECT_RESPONSE_TIMEOUT_MILLIS("tajo.network.connectresponsetimeoutmillis",
> 10000)
> NETWORK_RECEIVE_BUFFERSIZE("tajo.network.receivebuffersize", 1048576 * 10)
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
