[讨论]关于winsock10054和10053,10060的错误

flamen

因为我现在就碰到这个问题.查找了一下论坛,很多朋友都有碰到问题,但最后都是无果而终.虽然现在还没有解决问题,但已经有进展,共享出来,大家一起讨论.<br>
<br>
关于10054,10053,10060的错误<br>
<br>
10054:Wed 2005-07-13 10:43:46: Winsock Error 10054 Connection was reset by the other side!<br>
<br>
10053:Mon 2005-07-11 14:41:14: Winsock Error 10053 Software caused a connection abort.<br>
<br>
10060:Mon 2005-07-11 12:29:44: Winsock Error 10060 The connection timed out.<br>
<br>
MDaemon的官方解释<br>
<br>
WSAECONNABORTED (10053) Software caused connection abort <br>
<br>
A connection abort was caused internal to your host machine. The software caused a connection abort because there is no space on the socket's queue and the socket cannot receive further connections.<br>
<br>
WinSock description: The error can occur when the local network system aborts a connection. This would occur if WinSock aborts an established connection after data retransmission fails (receiver never acknowledges data sent on a datastream socket). <br>
<br>
TCP/IP scenario: A connection will timeout if the local system doesn't receive an (ACK)nowledgement for data sent. It would also timeout if a (FIN)ish TCP packet is not ACK'd (and even if the FIN is ACK'd, it will eventually timeout if a FIN is not returned). <br>
<br>
WSAECONNRESET (10054) Connection reset by peer <br>
<br>
A connection was forcibly closed by a peer. This normally results from a loss of the connection on the remote socket due to a timeout or a reboot. <br>
<br>
User suggestions: Some network systems have commands to report statistics. In this case, it might be possible to check the count of TCP RST packets received, or ICMP Port Unreachable packets. See other suggestions under WSAECONNABORTED. <br>
<br>
WSAETIMEDOUT (10060) Connection timed out <br>
<br>
A connect or send request failed because the connected party did not properly respond after a period of time. (The timeout period is dependent on the communication protocol.) <br>
<br>
Check the obvious first: check that the destination address is a valid IP address. If you used a hostname, did it resolve to the correct address? If the hostname resolution uses a local host table, it is possible you resolved to an obsolete address. Can you ping that hostname? <br>
<br>
Do you have a router configured? Is the router up and running? (You can check by pinging it, and then ping an address on the other side of it.) Try a traceroute to the destination address to check that all the routers are functioning. <br>
<br>
Check your subnet mask. If you don't have the proper subnet mask, your network system may treat a local address as a remote address (so it forwards addresses on the local subnet to the router, rather than broadcasting an ARP request locally), or vice versa. <br>
<br>
我还找到一个另外的解释<br>
Q.Why do I keep getting the message error Writing to Socket?<br>
A.If dial-up networking is uninstalled and then re-installed WITHOUT reapplying the NT service pack, MDaemon works fine.<br>
<br>
If you reapply SP4 (or SP5) like you should do after a DUN reinstall then you will instantly get the ''''Error writing to socket'''' when sending large messages. You then have to reinstall dial-up networking.<br>
<br>
我的环境是win 2003 SP1 + MD7.23英文版，放在win 2003 SP1 +ISA 2003 SP1 防火墙后．<br>
<br>
我的问题是一些大的邮件发不出去，而且不是所有的，有些可以发出．出现上面三个winsock错误.这些发不出的邮件，如果发不带附件的小邮件，就可以出去．<br>
<br>
开始时，我怀疑是ＤＮＳ的问题，测试后排除．想想也不应该是ＤＮＳ的问题，因为小邮件发送没问题．<br>
<br>
怀疑是不是ＩＳＡ有限制？解除ＩＳＡ所有过滤和连接数限制，也是不行．<br>
查看ＭＤ中的Active的Sessions，发现这些大邮件是开始是正常发送，过一会就停住，然后就出现10054或10060的错误．<br>
<br>
没办法，抓包看看．．发现下面的信息<br>
<br>
IP: .1.............. = Cannot fragment datagram<br>
<br>
<br>
搜索一下，发现微软的MS05-019的更新会引起这个问题，<br>
<a target=_blank href=http://support.microsoft.com/kb/898060/>http://support.microsoft.com/kb/898060/</a><br>
<br>
试着按微软的解决方法试，有效果，但问题还是没有彻底解决．<br>
<br>
但我想这应该是问题所在．现在只好等等微软的新补丁出来．<br>
<br>
不知道各位有没有这个问题，把解决过程发出来，大家一起分享一下．

observerzhou

<a target=_blank href=../Announce/Announce.asp?BoardID=34&ID=84017&p=3&q=1&r=28840>../Announce/Announce.asp?BoardID=34&ID=84017&p=3&q=1&r=28840</a><br>
<br>
LZ这是我以前发的帖子 这个问题的确讨厌<br>
<br>
我觉得MD的官方解释简直是P话。

flamen

现在我也不确定到底真正的原因是什么.是线路问题,还是数据包大小的问题,还是别的?<br>
<br>
observerzhou你的网络环境能描述一下吗?<br>
<br>
楼下的朋友如果有此同样问题,请把网络环境一起贴出.

goxia

本地内发送接受邮件测试，如果没有问题应该出在ISA上

flamen

不是所有的外发的大邮件会出现这种问题,只有几个包括公司总部,21cn等,发大附件时就会出错.而且在Active的All Sessions里查看,可以看到一开始发送时很正常,速度也很快,后面就突然停住了,然后就超时了.<br>
<br>
我现在已经没折了.....不知道要从哪下手,能试的都试过了

observerzhou

网通2M<br>
防火墙相关端口开放<br>
WIn2000 Ad server+MD 6.76<br>
在发到德国和某些日本的邮箱<br>
就会发生上述端口连接错误<br>
个人觉得是中国的国际出口太拥挤所至

jetsong

会不会是对方服务器的附件限制呢?<br>
一般附件不要超过2M为妙.

cuixinjun

可能是线路问题吧??但以前发送是正常的,后来以至到现在都不好用..服务器在美国???????????????///只是内部邮箱间发送有问题,用自己的邮箱往外部(像163等.)发邮件是没问题的.....挺烦心的...?????/求助,高手??

flamen

确认对方没有附件限制,本地也没有...现在用抓包分析中..

flamen

已经查明21cn的原因,从抓包看出21cn只能接收8M以下的邮件,所以不行.<br>
<br>
总部的原因是因为TCP previous segment lost....不知道是为什么会这样