×

Loading...
Ad by
  • 最优利率和cashback可以申请特批,好信用好收入offer更好。请点链接扫码加微信咨询,Scotiabank -- Nick Zhang 6478812600。
Ad by
  • 最优利率和cashback可以申请特批,好信用好收入offer更好。请点链接扫码加微信咨询,Scotiabank -- Nick Zhang 6478812600。

check this as well...

Another possibility is ident. Some servers use ident for reverse lookups. If ident is being dropped at the firewall, this will cause a delay. You can shorten this delay by adding a rule to reject ident (as opposed to drop), which will time ident out much quicker.

You can turn that off by modifying /etc/xinetd.d/telnet and commenting out the line that reads:

log_on_failure += USERID

If ident is the problem the telnet connections will establish quickly and you'll get a logon prompt after at most 30-40 seconds (DNS lookup delay). If that solves the problem we'll have to look at how to turn off ident for the other services.
Report

Replies, comments and Discussions:

  • 工作学习 / IT技术讨论 / 2个网络的怪问题
    1、一台服务器,如果PING,一点问题没有,很快,从TELNET过去开始,到出现LOGIN,要很长时间(大概5-10分钟),但出现LOGIN后又正常了

    2、另外一台,PING也是好的,但如果用FTP传文件,传到一部分后,就停止了,进程还在,但数据传输好象几乎停止了,尝试用RCP或者UUCP都一样。

    注:以上服务器都是LINUX系统
    • can't figure out anything form you post, but
      you can download a software called "ethreal" from gnu and intercept the packet form the network, you will be easy to figure out what happen
      • 我有空试试,你觉得大概是什么问题?
        • who knows? maybe something regarding TCP or TCP connection establishment
    • enable dns or edit /etc/host file.
      • 不觉得这两个文件有问题啊,站点地址都在HOSTS文件里面
        • the above suggestion deos makes sense. did you telnet ip address or machine name?? go ahead and try both
          • 应该试过了,那时候估计快疯了,啥都试过了
        • bear in mind, telnetd is trying reverse name lookup.
          • 双方的hosts文件都是完整的
            • make sure you using hosts before dns if you don't have dns setup correctly. (e.g nsswitch.conf)
    • this could help...
      This sounds like a reverse DNS lookup. Make entries in '/etc/hosts' for each machine. You should also have the line "order hosts, bind" in '/etc/host.conf' .

      check ifconfig for the info about RX errors and TX overruns, if they have large number of error, that could be a symptom of poor interrupt handling in the kernel of hardware.

      try use another NIC, see if the speed is improved...
      • check this as well...
        Another possibility is ident. Some servers use ident for reverse lookups. If ident is being dropped at the firewall, this will cause a delay. You can shorten this delay by adding a rule to reject ident (as opposed to drop), which will time ident out much quicker.

        You can turn that off by modifying /etc/xinetd.d/telnet and commenting out the line that reads:

        log_on_failure += USERID

        If ident is the problem the telnet connections will establish quickly and you'll get a logon prompt after at most 30-40 seconds (DNS lookup delay). If that solves the problem we'll have to look at how to turn off ident for the other services.