百度空间 | 百度首页 
 
查看文章
 
调整solaris系统增加邮件系统的性能[转]
2008-06-17 15:59
调整solaris系统增加邮件系统的性能


本文原来是专为sun的messaging server性能优化所写,虽然是非官方版本,但作者是sun的,相信也有一定权威性。
所以如果是用ims的朋友,直接看连接好了
http://ims.balius.com/downloads/iMS-Tuning-Guide.pdf
使用其他邮件系统但是环境再solaris下的,也可以参考其系统环境调整的部分,转贴如下(系统部分):
5. Solaris OE Considerations
5.1 TCP/IP Tuning
    Apply the following TCP/IP tunings to all mail servers.   These may
    also be appropriate for LDAP servers as well.   These values of these
    settings are geared towards high speed networks and lots of traffic
    to and from the servers.
    # ** Performance related **
    /usr/sbin/ndd -set /dev/tcp tcp_recv_hiwat 65536
    /usr/sbin/ndd -set /dev/tcp tcp_xmit_hiwat 65536
    /usr/sbin/ndd -set /dev/tcp tcp_conn_req_max_q 4096
    /usr/sbin/ndd -set /dev/tcp tcp_conn_req_max_q0 8192
    /usr/sbin/ndd -set /dev/tcp tcp_smallest_anon_port 8192
    /usr/sbin/ndd -set /dev/ip ip_ignore_redirect 1
    /usr/sbin/ndd -set /dev/tcp tcp_keepalive_interval 30000
    /usr/sbin/ndd -set /dev/tcp tcp_naglim_def 1
    # ** Security related **
    /usr/sbin/ndd -set /dev/tcp tcp_mss_min 108
    /usr/sbin/ndd -set /dev/ip ip_respond_to_echo_broadcast 0
    /usr/sbin/ndd -set /dev/ip ip_forward_directed_broadcasts 0
    /usr/sbin/ndd -set /dev/ip ip_strict_dst_multihoming 1
    /usr/sbin/ndd -set /dev/ip ip_forwarding 0
    /usr/sbin/ndd -set /dev/ip ip_forward_src_routed 0
    /usr/sbin/ndd -set /dev/ip ip_respond_to_timestamp 0
    /usr/sbin/ndd -set /dev/ip ip_respond_to_timestamp_broadcast 0
    #
    # Solaris guide says not to set lower than 60 seconds
    # should investigate further, but the following has worked
    /usr/sbin/ndd -set /dev/tcp tcp_time_wait_interval 15000
    #
    # Set according to local specifics.
    #/usr/sbin/ndd -set /dev/tcp tcp_mss_def 1460
5.2 /etc/system tuning
5.2.1 tcp_conn_hash_size
    set tcp_conn_hash_size=262144
5.2.2 File Descriptors
    # set hard limit on file descriptors
    set rlim_fd_max=4096
    # set soft limit on file descriptors

set rlim_fd_cur=4096
5.2.3 maxusers
set maxusers=2048
5.2.4 ncsize
Ideally you do not need to set this. Setting Section 5.2.3 to the
maximum value (2048) should allow the system to automatically tune
itself. You can use the following command
vmstat -s | grep cache
to see the percentage of hits against the Directory Name Lookup
Cache (DNLC). You want this to be as high as possible. If after
setting maxusers the cache hit rate against DNLC is not high enough
then setting ncsize is probably required.
ncsize = (4 * (max_nprocs + maxusers)) + 320
max_nprocs = 10 + (16 * maxusers)
maxusers = physmem - 2
5.2.5 Veritas Variables
If your system is using a Veritas file system you may need to adjust
two important variables.
vxfs:vxfs_ninode - VxFS inode structures held in memory
vxfs:vx_bc_bughwm - the Buffer CacheÕs Buffer High Water Mark
If your system uses Veritas Volume Manager (VxVM) then you should
look at vxio:vol_maxio. This variable controls the maximum size of
I/O requests that are sent down the SCSI chain without breaking the
request up. Recommendations are that this tunable parameter not
exceed 20% of kernel memory or physical memory (which ever is
smaller). This should match the size of your widest stripe.
5.2.6 Maximum Physical I/O Size
Increase maximum physical I/O size, the value shown here should work
for nearly all controllers.
set maxphys=8388608
You should also read http://www.samag.com/documents/s=7610/sam0210j/
0210j.htm as it covers this topic in depth.

类别:Opensolaris | 添加到搜藏 | 浏览() | 评论 (0)
 
最近读者:
 
网友评论:
发表评论:
姓 名:
网址或邮箱: (选填)
内 容:
验证码: 请点击后输入四位验证码,字母不区分大小写
      

     

©2009 Baidu