I think that, in general, a problem like that can be caused by "network drivers" (possibly virtual ones) that causes broadcast or other lengthy network iterations to be performed when someone tries to access the network.
The same software causes a similar problem with Rational Rose (10 minutes to load a simple model with InterDrive, 1 minute without Interdrive).
Ciao,
Danilo Turina
[EMAIL PROTECTED] wrote:
I just did a fresh install of 1.3.17 and setup telnetd. The login prompt is delayed by about 3 minutes! rlogin works just fine.This problem was described in detail in the mailing list towards the end of the summer. Is there a fix or a workaround? thanks/regards, Tom Rodman ------- Forwarded Message From: "Bruce Dobrin" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Subject: inetd telnet slowness Date: Fri, 6 Sep 2002 14:02:08 -0700 <snipped> I wrote about a month ago with this problem, but didn't manage to work out a viable solution. If I stop inetd and run "in.telnetd -debug" All is well. It appears to capture the machine/user correctly and I am imediately logged in (bypassing the login/passwd prompt). With inetd running, telnet starts, then hangs for about 3 minutes, finaly ginving me a "Login:" prompt. I assume that Inetd didn't pass usable machine/user information to the in.telnetd, and that the delay and eventual "login:" prompt is in.telnetd timing out waiting for login information, but I could be wrong about this. Rlogind works fine with inetd, and I assume it uses a similar validation mechanism. I'd appreciate any assistance. The diff of the Debug logs with and without inetd: with inetd: - -------------cut-------------- td: send suboption NEW-ENVIRON SEND td: send suboption TERMINAL-TYPE SEND td: ttloop td: netflush 249 chars td: ttloop read 73 chars td: netflush 26 chars td: recv suboption NAWS 0 80 (80) 0 25 (25) - -------------cut---------------- ........XON VARIABLE|ACK 17; XOFF VARIABLE|ACK 19; FORW1 VARIABLE|ACK 0; FORW2 VARIABLE|ACK 0; td: netflush 404 chars td: Entering processing loop ##### ##right here is where it hangs for 3 minutes} ##### CYGWIN_NT-5.0 1.3.12(0.54/3/2) (genevieve) (tty2) td: netflush 84 chars td: netread 9 chars td: recv do ECHO td: recv will BINARY td: recv wont LINEMODE td: netflush 85 chars login: td: netflush 7 chars - -------------cut---------------- Without inetd: - -------------cut---------------- td: send suboption NEW-ENVIRON SEND td: send suboption TERMINAL-TYPE SEND td: ttloop td: netflush 249 chars td: ttloop read 131 chars td: netflush 27 chars td: recv suboption NAWS 0 80 (80) 0 25 (25) - -------------cut---------------- .......XON VARIABLE|ACK 17; XOFF VARIABLE|ACK 19; FORW1 VARIABLE|ACK 0; FORW2 VARIABLE|ACK 0; td: netflush 404 chars td: Entering processing loop td: netflush 30 chars td: netread 9 chars td: recv do ECHO td: recv will BINARY td: recv wont LINEMODE td: netflush 85 chars Fanfare!!! You are successfully logged in to this server!!! - -------------cut---------------- <snipped> ------- End of Forwarded Message --v-v------------------C-U-T---H-E-R-E-------------------------v-v-- To: "James Garrison" <[EMAIL PROTECTED]> cc: <[EMAIL PROTECTED]> From: "Bruce Dobrin" <[EMAIL PROTECTED]> Subject: Re: Telnet takes minutes <mail headers snipped> -------------------------------------------------------------------- Thanks for the quick response. Those particular machines (1.3.12) are all on different segments from each other, with different os's (NT,XP, and 2K). But I have tried from other machines on both the same and different segments, same result, and from different telnet clients (SGI, Linux, NT, Cygwin...) it seems consistant. I even telneted to "localhost" and to the machine name and ip from itself.... same result. I assume reverse DNS is configured as it works (I didn't know it was configurable).. as a side note, rsh/rlogin works like a champ.... <snipped>
-- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Bug reporting: http://cygwin.com/bugs.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/