- 论坛徽章:
- 0
|
out of socket memory
通过此文章,我进行了相关调整,调整过后,就再未出现些相关提示,中文是我自己对照的翻译,水平有限!主要是调整TCP的缓冲区及对TCP连接状态进行调整及TCP状态队列
后两个之前有调整,这次只高速了TCP的缓冲区
Increasing the TCP send and receive buffers will increase the performance a lot if (and only if) you have a lot of large files to send.
增加TCP 发送和接收缓冲区对性能有很大的提高,在你有大文件发送的情况下
net.ipv4.tcp_wmem = 4096 65536 524288
net.core.wmem_max = 1048576
If you have a lot of large file uploads, increasing the receive buffers will help.
如果你有大文件上传,增大接收缓冲区对给你带来帮助- 01.net.ipv4.tcp_rmem = 4096 87380 524288
- 02.net.core.rmem_max = 1048576
- 复制代码
复制代码
- # These ensure that TIME_WAIT ports either get reused or closed fast.
- 以下参数确保TIME_WAIT 快速关闭
- 01.net.ipv4.tcp_fin_timeout = 1 ##打开fin_wait 快速关闭参数
- 02.net.ipv4.tcp_tw_recycle = 1 ##快速回收fin_wait参数
- 复制代码
复制代码 设置内核TCP的相关参数
# TCP memory- 01.net.core.rmem_max = 16777216 ##内核最大接收缓冲 16M
- 02.net.core.rmem_default = 16777216 ##内核?认接收缓冲 16M
- 03.net.core.netdev_max_backlog = 262144 ##2M
- 04.net.core.somaxconn = 262144 ##2M
- 复制代码
复制代码- 01.net.ipv4.tcp_syncookies = 1 ##打开SYN_COOKIE
- 02.net.ipv4.tcp_max_orphans = 262144 ##2M
- 03.net.ipv4.tcp_max_syn_backlog = 262144 ##SYN 队列大小 2M
- 04.net.ipv4.tcp_synack_retries = 2 ##SYN_ACK重试次数 2秒
- 05.net.ipv4.tcp_syn_retries = 2 ## SYN 重试次数 2秒
- 复制代码
复制代码 # you shouldn’t be using conntrack on a heavily loaded server anyway, but these are
# suitably high for our uses, insuring that if conntrack gets turned on, the box doesn’t die
如果用到IPTABLES 同样要增长ip_conntrack 的大小,这个是iptables 所管理的状态数目- 01.net.ipv4.ip_conntrack_max = 1048576
- 02.net.ipv4.netfilter.ip_conntrack_max = 1048576
- 复制代码
复制代码 |
|