版本:
客户端(client):
[root@desktop20 ~]# cat /proc/version
Linux version 2.6.32-220.el6.x86_64 ([email protected]) (gcc version 4.4.5 20110214 (Red Hat 4.4.5-6) (GCC) ) #1 SMP Wed Nov 9 08:03:13 EST 2011
[root@desktop20 ~]#
服务端(server):
[root@node1 ~]# cat /proc/version
Linux version 2.6.32-220.el6.x86_64 ([email protected]) (gcc version 4.4.5 20110214 (Red Hat 4.4.5-6) (GCC) ) #1 SMP Wed Nov 9 08:03:13 EST 2011
[root@node1 ~]#
症状:
在客户端执行ssh node1时,要等大概10秒才能出现输入密码的提示
解决:
================================================
使用strace跟踪,发现有个select函数用了10秒钟,查了半天也不知道到底是什么引起的(汗!-_- ...)
strace -o ssh.trc -fTttt ssh node4
view ssh.trc
... ...
520 38328 1367794302.169155 connect(4, {sa_family=AF_FILE, path="/tmp/keyring-24Tr0l/socket.ssh" }, 110) = 0 <0.000579>
521 38328 1367794302.169947 write(4, "\0\0\0\1", 4) = 4 <0.000040>
522 38328 1367794302.170091 write(4, "\v", 1) = 1 <0.000030>
523 38328 1367794302.170213 read(4, "\0\0\1<", 4) = 4 <0.001052>
524 38328 1367794302.171505 read(4, "\f\0\0\0\1\0\0\1\25\0\0\0\7ssh-rsa\0\0\0\1#\0\0\1\1\0\237n" ..., 316) = 316 <0.000041>
525 38328 1367794302.171831 write(3, "\177j\340\2263\3\245\26\360\231 \v\335]\364\353\276`t\256b .\3\204rc\360\364'yKT"..., 64) = 64 <0.000207>
526 38328 1367794302.172189 select(4, [3], NULL, NULL, NULL) = 1 (in [3]) <10.022289>
527 38328 1367794312.194794 read(3, "s\374\275q\361\\\333\306\30g\2K\216\362@\267LG~k\343\312\23 2\223lJ\206\32s\330%H"..., 8192) = 80 <0.000014>
528 38328 1367794312.194905 write(3, "=\310\33\227X\3142=R&*\302a\320\0R\202\334\372)\215\24\264 \20\256\tZ\20\332\263\357\245"..., 368) = 368 <0.000075>
529 38328 1367794312.195019 select(4, [3], NULL, NULL, NULL) = 1 (in [3]) <0.007280>
530 38328 1367794312.202396 read(3, "\261\200\311L\240\342\332\213:J\246\t\7Ctk\224\16\362\370\2 7\275\33^f\327D\3504\337\323\344"..., 8192) = 320 <0.000026>
531 38328 1367794312.202557 write(4, "\0\0\2\223", 4) = 4 <0.000038>
... ...
526: vim里的行号
38328:进程ID
1367794302.172189:自1970-01-01 00:00:00 UTC到现在的秒数,精确到小数点后六位
select(4, [3], NULL, NULL, NULL) :select函数和参数
= 1:select函数的返回值
(in [3]) :?
<10.022289> :select函数用了多少秒,,精确到小数点后六位
================================================
strace搞不定,试试开启ssh的debug信息,看看能查到什么
ssh -vvv node1
... ...
debug1: Unspecified GSS failure. Minor code may provide more information
Credentials cache file '/tmp/krb5cc_0' not found
debug1: Unspecified GSS failure. Minor code may provide more information
... ...
看来是客户端启用了kerbose验证,但是我没打算使用kerbose验证,于是关了这个选项。
[root@desktop20 ssh]# grep GSSAPIAuthentication /etc/ssh/ssh_config
GSSAPIAuthentication no
[root@desktop20 ssh]#
关了之后,之前的提示没有了,可是问题依旧,ssh node1还是要等10秒。
(事实上GSS验证超时确实是发起ssh连接后要等很长时间才能输入密码的一个原因,但在这篇文章里只是原因之一,所以这里关闭客户端的GSS验证效果并不明显)
再执行ssh -vvv node1时发现总会在“debug3: Wrote 64 bytes for a total of 1109”卡上一小段时间
... ...
debug1: SSH2_MSG_NEWKEYS received
debug1: SSH2_MSG_SERVICE_REQUEST sent
debug3: Wrote 48 bytes for a total of 1045
debug2: service_accept: ssh-userauth
debug1: SSH2_MSG_SERVICE_ACCEPT received
debug2: key: /root/.ssh/identity ((nil))
debug2: key: /root/.ssh/id_rsa (0x7f1391a6d970)
debug2: key: /root/.ssh/id_dsa ((nil))
debug3: Wrote 64 bytes for a total of 1109
debug1: Authentications that can continue: publickey,gssapi-keyex,gssapi-with-mic,password
debug3: start over, passed a different list publickey,gssapi-keyex,gssapi-with-mic,password
debug3: preferred publickey,keyboard-interactive,password
debug3: authmethod_lookup publickey
debug3: remaining preferred: keyboard-interactive,password
... ...
================================================
GOOGLE一圈之后发现不少回答都提到了服务端打开的UseDNS选项可能是问题的原因,于是关闭服务端UseDNS选项,重启sshd服务
[root@node1 ~]# grep UseDNS /etc/ssh/sshd_config
UseDNS no
[root@node1 ~]# /etc/init.d/sshd restart
Stopping sshd: [ OK ]
Starting sshd: [ OK ]
[root@node1 ~]#
果然,现在从客户端ssh node1几乎是秒连。
REF:
1. ssh 连接很慢的解决办法
http://blog.csdn.net/ablo_zhou/article/details/5074887
2. Slow SSH authentication times
http://ubuntuforums.org/showthread.php?t=1605241