- stunnel-users - stunnel.org

Re-2: [stunnel-users] Problem with mysql
by lmarcilly＠aressi.fr 20 Jul '05

20 Jul '05

I have just remember that tcp_wrapper is not installed on my box... When compiling stunnel, i use : ./configure --prefix=/usr --disable-libwrap make make install I think libwrap is not necessary but perhaps i am wrong... Tcp wrapper is used in order to filter traffic? Could you explain me? Thanks a lot! -------- Original Message -------- Subject: Re: [stunnel-users] Problem with mysql (20-juil.-2005 14:47) From: lmarcilly(a)aressi.fr To: lmarcilly(a)aressi.fr > I try to analyse traffic with tcpdump and when i use stunnel, there is no > packets with my linux box ip so i think stunnel doesn't work properly... > Here is the stunnel.log : > > 2005.07.20 12:58:23 LOG5[14714:1]: stunnel 4.11 on i686-pc-linux-gnu > UCONTEXT+POLL+IPv4 with OpenSSL 0.9.7g 11 Apr 2005 > 2005.07.20 12:58:23 LOG7[14714:1]: Snagged 64 random bytes from /root/.rnd > 2005.07.20 12:58:23 LOG7[14714:1]: Wrote 1024 new random bytes to /root/. > rnd > 2005.07.20 12:58:23 LOG7[14714:1]: RAND_status claims sufficient entropy > for the PRNG > 2005.07.20 12:58:23 LOG6[14714:1]: PRNG seeded successfully > 2005.07.20 12:58:23 LOG7[14714:1]: Certificate: /usr/etc/stunnel/stunnel. > pem > 2005.07.20 12:58:23 LOG7[14714:1]: Key file: /usr/etc/stunnel/stunnel.pem > 2005.07.20 12:58:23 LOG6[14714:1]: file ulimit = 1024 (can be changed with ' > ulimit -n') > 2005.07.20 12:58:23 LOG6[14714:1]: poll() used - no FD_SETSIZE limit for > file descriptors > 2005.07.20 12:58:23 LOG5[14714:1]: 500 clients allowed > 2005.07.20 12:58:23 LOG7[14714:1]: FD 4 in non-blocking mode > 2005.07.20 12:58:23 LOG7[14714:1]: FD 5 in non-blocking mode > 2005.07.20 12:58:23 LOG7[14714:1]: FD 6 in non-blocking mode > 2005.07.20 12:58:23 LOG7[14714:1]: SO_REUSEADDR option set on accept socket > 2005.07.20 12:58:23 LOG7[14714:1]: mysql bound to 0.0.0.0:3306 > 2005.07.20 12:58:23 LOG7[14715:1]: Created pid file /stunnel.pid > 2005.07.20 14:07:13 LOG5[14715:0]: Received signal 15; terminating > 2005.07.20 14:07:13 LOG7[14715:0]: removing pid file /stunnel.pid > > When i look at stunnel.pid, there is 231 in the file and it is the stunnel > pid... > > In my stunnel.conf, i have put client=yes and > > [mysql] > accept = 3306 > connect = 192.198.10.120:3307 > > and on my stunnel server : > > [mysql] > accept = 3307 > connect = localhost:3306 > > It works with my mysql client on windows (v4.1.12a) but not with my mysql > client on linux (v4.1.8a). Perhaps the difference of version is the problem > but i think not... > > Which options are applied when windows binaries of stunnel are compiled? > Perhaps i have to change option when i compile on my linux box... > > Thanks in advance! > > lm. > > -------- Original Message -------- > Subject: [stunnel-users] Problem with mysql (20-juil.-2005 11:41) > From: lmarcilly(a)aressi.fr > To: lmarcilly(a)aressi.fr > > > Hi all, > > > > i try to encrypt mysql connections... > > > > I saw this : http://www.stunnel.org/examples/mysql.html > > > > but i use stunnel4.x. > > > > I configure it on my windows box and it work well! (mysql server is on a > > linux box, i have configure stunnel as a server on this box) > > > > when i try mysql -u user --password=xyz it works > > > > but i have to made it work on a linux box. > > I use the same config file for stunnel, but when i try > > > > mysql -u user --password=xyz i have : > > Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld. > > sock' > > > > This socket doesn't exist since there is no MySQL server on the box. > > > > I try with --protocol=TCP but it won't work, i have : > > > > Can't connect to MySQL server on 'localhost' > > > > Do you have an idea? Do you want more details on my config file for stunnel? > > > > > > > > Thanks in advance and sorry for my english! > > > > lm > > > > > > _______________________________________________ > > stunnel-users mailing list > > stunnel-users(a)mirt.net > > http://stunnel.mirt.net/mailman/listinfo/stunnel-users > > > To: lmarcilly(a)aressi.fr > stunnel-users(a)mirt.net To: lmarcilly(a)aressi.fr stunnel-users(a)mirt.net

1 0

Problem with mysql
by lmarcilly＠aressi.fr 20 Jul '05

20 Jul '05

Hi all, i try to encrypt mysql connections... I saw this : http://www.stunnel.org/examples/mysql.html but i use stunnel4.x. I configure it on my windows box and it work well! (mysql server is on a linux box, i have configure stunnel as a server on this box) when i try mysql -u user --password=xyz it works but i have to made it work on a linux box. I use the same config file for stunnel, but when i try mysql -u user --password=xyz i have : Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' This socket doesn't exist since there is no MySQL server on the box. I try with --protocol=TCP but it won't work, i have : Can't connect to MySQL server on 'localhost' Do you have an idea? Do you want more details on my config file for stunnel? Thanks in advance and sorry for my english! lm

1 0

SSL Session Cache and HTTPS performance
by Daniel Hamburg 19 Jul '05

19 Jul '05

Hello everybody, I hope I didn't muss an existing thread about this question, but I wasn't able to find one. At least none where I found an answer, although my question is quite simple. I would like to test the performance difference between enabled and disabled SSLCaching. I use a configuration file for Stunnel 4.11 (daemon mode) where I tell it to accept incoming connections on port 444 and direct them to a local web-server on port 80. Because STunnel uses Caching by default, I tried to tell STunnel to disable SSL Caching by setting session = 0 in the configuration file, which resulted in an error when starting the program. I looked at the code and found that there is a check so that the session parameter can not be equal to 0. I disabled the check and started stunnel again which seemed to work. But taking a look at the packets exchanged via a browser and the server resulted in the conclusion that caching is still enabled. My simple question is, how I can disable the Caching. Another question is if anyone of you has made any performance tests regarding STunnel in combination with an Apache HTTP server on the same machine. I'm specially interested in modes to get this combination run faster. And now please do not tell me to use mod_ssl. :) Thanks in advance. Daniel -- ====================================================================== Dipl.-Ing. Daniel Hamburg Tel.: 0234/32-28034 Ruhr-Universität Bochum Fax.: 0234/32-08034 Arbeitsgruppe Integrierte Informationssysteme IC 4/47 Universitätsstraße 150 Web: www.iis.ruhr-uni-bochum.de D-44780 Bochum eMail: daniel.hamburg(a)iis.rub.de ======================================================================

3 5

smtps problem on Fedora Core 4
by Anders Gjerløv 19 Jul '05

19 Jul '05

Hi, I am trying to get the smtps solution described on http://www.linuxquestions.org/questions/answers.php?action=viewarticle&arti… working on my server. Basically I want my users on Windows XP boxes to get their mail from 127.0.0.1:100 and send their mail to 127.0.0.1:25 I have gotten as far as having the users able to get mails. That is I have stunnel on a windows box, with a configuration file like this: client = yes debug = debug cert = pemfile.pem [pop3s] accept = localhost:110 connect = my.domain.com:995 [smtps] accept = localhost:25 connect = my.domain.com:465 And a telnet to port 110 on the lookback interface gives me access to the pop server. Så I am guessing that the certificates etc. are working correctly. As for the smtps I am using xinetd, and have tried with various configurations. Here is one of the one's that didn't work: service smtps { disable = no socket_type = stream wait = no user = root server = /usr/sbin/stunnel server_args = -v3 -rlocalhost:25 log_on_success += HOST DURATION log_on_failure += HOST ATTEMPT } Any ideas as to how to setup my xinetd file (or is there a better solution?) # stunnel -version stunnel 4.08 on i386-redhat-linux-gnu PTHREAD+POLL+IPv4+LIBWRAP with OpenSSL 0.9.7f 22 Mar 2005 Global options cert = /etc/stunnel/stunnel.pem ciphers = ALL:!ADH:+RC4:@STRENGTH debug = 5 key = /etc/stunnel/stunnel.pem pid = /var/run/stunnel.pid RNDbytes = 64 RNDfile = /dev/urandom RNDoverwrite = yes session = 300 seconds verify = none Service-level options TIMEOUTbusy = 300 seconds TIMEOUTclose = 60 seconds TIMEOUTconnect = 10 seconds TIMEOUTidle = 43200 seconds Kind regards, Anders

2 1

Re: [stunnel-users] HTTPS Proxy support?
by Sean Covel 17 Jul '05

17 Jul '05

Yeah, good for you. Don't we all. I'll just go shove off then. Sean Michal Trojnara wrote: > Sean, > >> $25. > > > It's not enough to allocate my time. 8-) > I usually work for $75/hour. > > Anyway I'm going to support this feature some day. > > Best regards, > Mike > > Sean Covel wrote: > Mike, > > $25. > > I have a paypal link on my development site. Rarely see a payment, but > I have seen some. Doesn't cost anything to have the link or the paypal > account. > > Sean > > Michal Trojnara wrote: > >>Sean, >> >> >>>Do you have a PayPal link somewhere? >> >> >>So far I didn't use PayPal. >>How much money would you like to spend to sponsor stunnel development? >> >>Best regards, >> Mike >> > >

1 0

Stunnel on openwrt
by Markus Mauch 15 Jul '05

15 Jul '05

Hello, I'm using stunnel on a asus router with a openwrt like linux on it. The Idea is to secure traffic from and to a thttpd web server which is running on that box as well. Problem is that stunnel only seems to be able to handle one connection at a time. Whenever it has to cope with more than one it crashes. Unfortunatelly this happens all the time because usually a web browser opens several tcp connections in order to speed things up. So whenever I browse to a web page on the router through stunnel I get only the text. The rest (e.g. the images) which is loaded via different tcp connections does not appear and stunnel won't respond to any further requests until it is restarted. My explanation to this is that the very first request is processed normally but as soon as one other connection causes stunnel to create a second process the whole thing goes to the dogs. ps reveals that there are always two instances of stunnel running after such an experiment, both hanging, of course. So my question is what could I do to keep stunnel from creating new processes when a client tries to communicate over more than one tcp connection or - better - what chould I do in the first place to prevent stunnel from hanging in such a case? Any help would be very apreciated...

1 0

Re: [stunnel-users] HTTPS Proxy support?
by seanecovel＠comcast.net 15 Jul '05

15 Jul '05

Mike, GREAT! I did see a patch out on the web that seems to add this feature. Any chance that patch will be included in MAIN? I don't want to have to patch stunnel on multiple OS especially when you are going to implement the feature yourself. What can I do to "encourage" you to add that support sooner than later? Sean -------------- Original message -------------- > "Sean E. Covel" wrote: > > Does stunnel support HTTPS Proxies with authentication? > > Currently not. > > > Could it? > > Yes, CONNECT protocol support (RFC 2817 section 5.2) is on my TODO list: > http://stunnel.mirt.net/todo_sdf.html > > Best regards, > Mike > _______________________________________________ > stunnel-users mailing list > stunnel-users(a)mirt.net > http://stunnel.mirt.net/mailman/listinfo/stunnel-users

3 2

HTTPS Proxy support?
by Sean E. Covel 14 Jul '05

14 Jul '05

I am trying to use stunnel to tunnel VNC through a firewall using port 443. When I test on 2 machines on the LAN everything works fine. Nice tool! When I try to connect through the firewall I suspect the HTTPS Proxy is getting in the way. It requires authentication. Does stunnel support HTTPS Proxies with authentication? Could it? Thanks, Sean

2 1

Q: controlled access to service?
by Nardmann, Heiko 14 Jul '05

14 Jul '05

Hi, I want to control access to a through stunnel reachable service. Only those clients shall be allowed to use the service which provide a known certificate. I have found the option "CApath"; can this directory be used to collect all client certificates? Or is it absolutely necessary to have CA certs there? Another thing in this environment: I do not know or own every CA certificate used by the clients - I only get the client certificates itself. So I want to do only a one-level client cert verification. Which verify level do I need for this? 2 or 3? What about removing certificates from the CApath directory? Do I have to restart stunnel to make this change be effective? Another thing: since the client certificates are not revoked by us I am not able to use CRLs for controlling access to our service. -- Heiko Nardmann (Dipl.-Ing. Technische Informatik) secunet Security Networks AG - Sicherheit in Netzwerken (www.secunet.de), Weidenauer Str. 223-225, D-57076 Siegen Tel. : +49 271 48950-13, Fax : +49 271 48950-50

3 2

Re: [stunnel-users] Certificate renaming
by Paul Jones 13 Jul '05

13 Jul '05

>"The easiest option is to use CAfile instead of CApath." So does the CA (Certificate Authority) certificate still go in the CAfile? That would mean the CAfile contains the CA certificate and all client certificates (on the server side), right? Cheers, Paul. _________________________________________________________________ REALESTATE: biggest buy/rent/share listings http://ninemsn.realestate.com.au

1 0

debugging stunnel
by edonald1＠rochester.rr.com 12 Jul '05

12 Jul '05

Been trying to debug stunnel issues for stunnel-4.09 on Solaris 2.8. When in trouble I resort to reading documentation. It is now that I notice these two small lines in the begining of the document called INSTALL: 1. If your machine supports POSIX threads make sure your SSL library is compiled with -DTHREADS. I did not do this .... ( verify openssl is compiled with -DTHREADS). WHat is the result of not compiling openssl with THREADS? Also what version of openssl is recommended. I am using OpenSSL 0.9.6m . Thanks in advance, Liz

2 1

Solaris 2.8 stunnel stop running
by edonald1＠rochester.rr.com 12 Jul '05

12 Jul '05

Running stunnel on Solaris 2.8. Stunnel running as client quits unexpectedly . here are the lines in the log: 2005.07.12 11:01:31 LOG7[5736:1]: mod_jk_secure accepted FD=2 from 127.0.0.1:57610 2005.07.12 11:01:31 LOG7[5736:1]: FD 2 in non-blocking mode 2005.07.12 11:02:28 LOG3[5736:4]: SSL_connect: Peer suddenly disconnected 2005.07.12 11:02:28 LOG7[5736:4]: linger (remote): Invalid argument (22) 2005.07.12 11:02:28 LOG7[5736:4]: mod_jk_secure finished (1 left) 2005.07.12 11:02:28 LOG7[5736:5]: mod_jk_secure started 2005.07.12 11:02:28 LOG7[5736:5]: TCP_NODELAY option set on local socket 2005.07.12 11:02:28 LOG5[5736:5]: mod_jk_secure connected from 127.0.0.1:57610 2005.07.12 11:02:28 LOG7[5736:5]: FD 0 in non-blocking mode 2005.07.12 11:02:28 LOG7[5736:5]: mod_jk_secure connecting 150.247.130.5:51001 2005.07.12 11:02:28 LOG7[5736:5]: connect_wait: waiting 10 seconds 2005.07.12 11:02:28 LOG7[5736:1]: mod_jk_secure accepted FD=1 from 127.0.0.1:57614 2005.07.12 11:02:28 LOG7[5736:1]: FD 1 in non-blocking mode 2005.07.12 11:02:29 LOG7[5736:6]: mod_jk_secure started 2005.07.12 11:02:29 LOG7[5736:6]: TCP_NODELAY option set on local socket 2005.07.12 11:02:29 LOG5[5736:6]: mod_jk_secure connected from 127.0.0.1:57614 2005.07.12 11:02:29 LOG7[5736:6]: FD 7 in non-blocking mode 2005.07.12 11:02:29 LOG7[5736:6]: mod_jk_secure connecting 150.247.130.5:51001 2005.07.12 11:02:29 LOG7[5736:6]: connect_wait: waiting 10 seconds 2005.07.12 11:02:29 LOG7[5736:5]: connect_wait: connected 2005.07.12 11:02:29 LOG7[5736:5]: Remote FD=0 initialized 2005.07.12 11:02:29 LOG7[5736:5]: TCP_NODELAY option set on remote socket 2005.07.12 11:02:29 LOG7[5736:5]: SSL state (connect): before/connect initialization 2005.07.12 11:02:29 LOG7[5736:5]: SSL state (connect): SSLv3 write client hello A 2005.07.12 11:02:29 LOG7[5736:6]: connect_wait: connected 2005.07.12 11:02:29 LOG7[5736:6]: Remote FD=7 initialized 2005.07.12 11:02:29 LOG7[5736:6]: TCP_NODELAY option set on remote socket 2005.07.12 11:02:29 LOG7[5736:6]: SSL state (connect): before/connect initialization 2005.07.12 11:02:29 LOG7[5736:6]: SSL state (connect): SSLv3 write client hello A All ideas welcome. I happen to know that the server side stunnel is down. SHouldn't the client side keep running under these circumstances? Thanks, Liz

1 0

stunnel 4.11 relaesed
by Michal Trojnara 12 Jul '05

12 Jul '05

Dear Users, Here is the ChangeLog entry: Version 4.11, 2005.07.09, urgency: MEDIUM: * New features - New ./configure option --with-threads to select thread model. - ./configure option --with-tcp-wrappers renamed to --disable-libwrap. I hope the meaning of the option is much more clear, now. * Bugfixes - Workaround for non-standard makecontext() uc_stack.ss_sp parameter semantics on Sparc/Solaris 9 and earlier. - scan_waiting_queue() no longer drops contexts. - Inetd mode coredumps with UCONTEXT fixed. - Cleanup context is no longer used. - Releasing memory of the current context is delayed. - Win32 headers reordered for Visual Studio 7. - Some Solaris compilation warnings fixed. - Rejected inetd mode without 'connect' or 'exec'. * Release notes - UCONTEXT threading seems stable, now. Upgrade is recommended. Home page, downloads: http://stunnel.mirt.net/ sha1sum for stunnel-4.11.tar.gz file: cf57169d591fbe3371a29e432d840e7f66103a9f Best regards, Mike

3 3

Stunnel-4.10 and Stunnel-4.11 in HPUX-11.23 IPF gives coredump
by Punitha 12 Jul '05

12 Jul '05

Hello All, I'm using HP-UX11.23 IPF. I have built Stunnel-4.11 with OpenSSL - 0.9.7e. I could build it successfully but I'm failed in connecting with telnet. After a while connecting to the telnet the client gets coredump and kills the process. So the connection is no more established. Note: I had the same problem in Stunnel-4.10 also but not before that. I could very well compile and use Stunnel-4.09. My configuration file for client is: cert = /opt/iexpress/stunnel/etc/stunnel/stunnel.pem chroot = /opt/iexpress/stunnel/var/stunnel pid = setuid = root setgid = root CAfile = /opt/iexpress/stunnel/etc/stunnel/stunnel.pem debug = 7 client = yes foreground = yes verify = 3 [test1] accept = 8080 connect = 9090 For server: cert = /opt/iexpress/stunnel/etc/stunnel/stunnel.pem chroot = /opt/iexpress/stunnel/var/stunnel pid = setuid = root setgid = root CAfile = /opt/iexpress/stunnel/etc/stunnel/stunnel.pem debug = 6 client = no foreground = yes verify = 1 [test1] accept = 9090 connect = 23 The client log shows as follows: # ../../sbin/stunnel stunnel-client.conf 2005.07.10 23:39:26 LOG5[8909:1]: stunnel 4.11 on ia64-hp-hpux11.23 UCONTEXT+POLL+IPv4+LIBWRAP with OpenSSL 0.9.7e 25 Oct 2004 2005.07.10 23:39:26 LOG7[8909:1]: Snagged 64 random bytes from /dev/random 2005.07.10 23:39:26 LOG7[8909:1]: RAND_status claims sufficient entropy for the PRNG 2005.07.10 23:39:26 LOG6[8909:1]: PRNG seeded successfully 2005.07.10 23:39:26 LOG7[8909:1]: Certificate: /home/punitha/stunnel1/etc/stunnel/stunnel.pem 2005.07.10 23:39:26 LOG7[8909:1]: Key file: /home/punitha/stunnel1/etc/stunnel/stunnel.pem 2005.07.10 23:39:26 LOG7[8909:1]: Loaded verify certificates from /home/punitha/stunnel1/etc/stunnel/stunnel.pem 2005.07.10 23:39:26 LOG6[8909:1]: file ulimit = 2048 (can be changed with 'ulimit -n') 2005.07.10 23:39:26 LOG6[8909:1]: poll() used - no FD_SETSIZE limit for file descriptors 2005.07.10 23:39:26 LOG5[8909:1]: 1000 clients allowed 2005.07.10 23:39:26 LOG7[8909:1]: FD 3 in non-blocking mode 2005.07.10 23:39:26 LOG7[8909:1]: FD 4 in non-blocking mode 2005.07.10 23:39:26 LOG7[8909:1]: FD 5 in non-blocking mode 2005.07.10 23:39:26 LOG7[8909:1]: SO_REUSEADDR option set on accept socket 2005.07.10 23:39:26 LOG7[8909:1]: test1 bound to 0.0.0.0:8080 2005.07.10 23:39:26 LOG7[8909:1]: No pid file being created 2005.07.10 23:40:08 LOG7[8909:1]: test1 accepted FD=8 from 127.0.0.1:49700 2005.07.10 23:40:08 LOG7[8909:1]: Creating a new context 2005.07.10 23:40:08 LOG7[8909:1]: Context 2 created 2005.07.10 23:40:08 LOG7[8909:2]: Context swap: 1 -> 2 Memory fault(coredump) # It will be helpful if anyone of you guide me in this. Thanx In Advance. Regards, Punitha, V.

2 3

Re: [stunnel-users] Poll/Descriptor Problem
by Andrew Gay 11 Jul '05

11 Jul '05

Hi I get the same/similar error messages using Stunnel 4.10 (client) on Solaris 9 (sparc). I also see some very large integer numbers (e.g. FD=2615261238) in the log which art file descriptors. Perhaps there is an issue with big/little endedness in the later Stunnel releases. I went back to using Stunnel 4.07 and the problems went away. Andrew Gay > > I am running Stunnel 4.09 on Mac OS X 10.4 to provide HTTPS service on > the standard ports (accept on 443, connect to 80). > > All connections are either truncated or dropped entirely, with the > following messages written to the log: > > ------------------------------------------ > https connected from ... > INTERNAL ERROR:s_poll_wait returned 1, but no descriptor is ready > Connection reset: XXXXXX bytes sent to SSL, XXX bytes sent to socket > ------------------------------------------ > > I've tried 4.10, but the result is exactly the same. > > Does anybody have any thoughts on what might be going on? > > Thanks, > John >

2 1

Poll/Descriptor Problem
by John O'Fallon 08 Jul '05

08 Jul '05

I am running Stunnel 4.09 on Mac OS X 10.4 to provide HTTPS service on the standard ports (accept on 443, connect to 80). All connections are either truncated or dropped entirely, with the following messages written to the log: ------------------------------------------ https connected from ... INTERNAL ERROR:s_poll_wait returned 1, but no descriptor is ready Connection reset: XXXXXX bytes sent to SSL, XXX bytes sent to socket ------------------------------------------ I've tried 4.10, but the result is exactly the same. Does anybody have any thoughts on what might be going on? Thanks, John

1 0

Re: [stunnel-users] Certificate renaming
by Paul Jones 07 Jul '05

07 Jul '05

>From the discussion here, I got the impression that it is still possible for Stunnel to run correctly and successful connections to be made if the certificate files are NOT renamed using the xxxxx.0 format, where xxxxx is the hash value of the certificate. So I simply placed the .crt files for all clients on the server (under the directory specified in the CApath field in the CONF file). Now, I've tried, tried, and tried, but it still doesn't work. I get a "bad certificate" error when attempting to connect from the client. Am I missing something? Note: I am running Stunnel on Win2K. Cheers. Paul _________________________________________________________________ Single? Start dating at Lavalife. Try our 7 day FREE trial! http://lavalife9.ninemsn.com.au/clickthru/clickthru.act?context=an99&locale…

1 0

Re: [stunnel-users] Stunnel for pop3 on solaris 2.6 (more info)
by Douglas Phillipson 06 Jul '05

06 Jul '05

Should I have "protocol = pop3" in my config file? When I do, I get this from the stunnel log: 2005.07.06 12:46:54 LOG7[18045:0]: CONTEXT 1, FD=6, (IN)->() 2005.07.06 12:46:54 LOG7[18045:0]: CONTEXT 2, FD=0, (IN)->(IN) 2005.07.06 12:46:54 LOG7[18045:0]: Waiting 300 second(s) for 3 file descriptor(s) 2005.07.06 12:46:54 LOG7[18045:0]: CONTEXT 1, FD=4, (IN)->() 2005.07.06 12:46:54 LOG7[18045:0]: CONTEXT 1, FD=6, (IN)->() 2005.07.06 12:46:54 LOG7[18045:0]: CONTEXT 2, FD=0, (IN)->(IN) 2005.07.06 12:46:54 LOG7[18045:2]: <- .g.. 2005.07.06 12:46:54 LOG3[18045:2]: Client does not want TLS 2005.07.06 12:46:54 LOG5[18045:2]: Protocol negotiation failed 2005.07.06 12:46:54 LOG3[18045:2]: Protocol negotiations failed 2005.07.06 12:46:54 LOG7[18045:2]: pop3s finished (0 left) 2005.07.06 12:46:54 LOG7[18045:2]: Context 2 closed 2005.07.06 12:46:54 LOG7[18045:0]: Waiting -1 second(s) for 2 file descriptor(s) Doug P Douglas Phillipson wrote: > I'm not sure it it applies but if I do: > > openssl s_client -connect 172.20.12.59:995 > > I get the following error: > > CONNECTED(00000003) > 17964:error:140790E5:SSL routines:SSL23_WRITE:ssl handshake failure:s23_lib.c:226: > > > Doug P > > > Douglas Phillipson wrote: > >> I have a Solaris 2.6 box and am trying to get pop3 over SSL running. I use: >> >> qpopper 4.0.5 >> openssl-0.9.7g >> stunnel 4.10 compiled with gcc 2.95.3 >> >> When I compiled stunnel it made a private ket and certificate in /usr/local/etc/stunnel/stunnel.pem. >> >> Do I need anything else? >> >> I have the following configuration: >> >> Inetd.conf: >> >> pop3 stream tcp nowait root /usr/local/sbin/popper qpopper -S -t /poplog >> >> stunnel.conf: >> >> cert = /usr/local/etc/stunnel/stunnel.pem >> key = /usr/local/etc/stunnel/stunnel.pem >> debug = 7 >> output = /stunnel.log >> pid = /stunnel.pid >> client = yes >> >> [pop3s] >> accept = 995 >> connect = 110 >> >> I run stunnel and get the following output: >> >> 2005.07.06 11:34:17 LOG5[17873:1]: stunnel 4.10 on sparc-sun-solaris2.6 UCONTEXT+POLL+IPv4 with OpenSSL 0.9.7g 11 Apr 2005 >> 2005.07.06 11:34:18 LOG7[17873:1]: Snagged 64 random bytes from //.rnd >> 2005.07.06 11:34:18 LOG7[17873:1]: Wrote 1024 new random bytes to //.rnd >> 2005.07.06 11:34:18 LOG7[17873:1]: RAND_status claims sufficient entropy for the PRNG >> 2005.07.06 11:34:18 LOG6[17873:1]: PRNG seeded successfully >> 2005.07.06 11:34:18 LOG7[17873:1]: Certificate: /usr/local/etc/stunnel/stunnel.pem >> 2005.07.06 11:34:18 LOG7[17873:1]: Key file: /usr/local/etc/stunnel/stunnel.pem >> 2005.07.06 11:34:18 LOG6[17873:1]: file ulimit = 64 (can be changed with 'ulimit -n') >> 2005.07.06 11:34:18 LOG6[17873:1]: poll() used - no FD_SETSIZE limit for file descriptors >> 2005.07.06 11:34:18 LOG5[17873:1]: 29 clients allowed >> 2005.07.06 11:34:18 LOG7[17873:1]: FD 4 in non-blocking mode >> 2005.07.06 11:34:18 LOG7[17873:1]: FD 5 in non-blocking mode >> 2005.07.06 11:34:18 LOG7[17873:1]: FD 6 in non-blocking mode >> 2005.07.06 11:34:18 LOG7[17873:1]: SO_REUSEADDR option set on accept socket >> 2005.07.06 11:34:18 LOG7[17873:1]: pop3s bound to 0.0.0.0:995 >> 2005.07.06 11:34:18 LOG7[17874:1]: Created pid file /stunnel.pid >> 2005.07.06 11:34:18 LOG7[17874:0]: Waiting -1 second(s) for 2 file descriptor(s) >> >> >> I connect via pop3 in thunderbird with ssl and qpopper always says: >> (null) at localhost (127.0.0.1): -ERR Unknown command: "". >> (nulI/O error flushing output to client at localhost [127.0.0.1]: Broken pipe (32)l) at localhost (127.0.0.1): -ERR POP EOF or I/O Error >> >> Stunnel says: >> >> 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=4, (IN)->() >> 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=6, (IN)->(IN) >> 2005.07.06 11:37:13 LOG7[17884:1]: pop3s accepted FD=0 from 172.20.10.7:45464 >> 2005.07.06 11:37:13 LOG7[17884:1]: Creating a new context >> 2005.07.06 11:37:13 LOG7[17884:1]: Context 2 created >> 2005.07.06 11:37:13 LOG7[17884:2]: pop3s started >> 2005.07.06 11:37:13 LOG7[17884:2]: FD 0 in non-blocking mode >> 2005.07.06 11:37:13 LOG7[17884:2]: TCP_NODELAY option set on local socket >> 2005.07.06 11:37:13 LOG5[17884:2]: pop3s connected from 172.20.10.7:45464 >> 2005.07.06 11:37:13 LOG7[17884:2]: FD 1 in non-blocking mode >> 2005.07.06 11:37:13 LOG7[17884:2]: pop3s connecting 127.0.0.1:110 >> 2005.07.06 11:37:13 LOG7[17884:2]: Remote FD=1 initialized >> 2005.07.06 11:37:13 LOG7[17884:2]: TCP_NODELAY option set on remote socket >> 2005.07.06 11:37:13 LOG7[17884:2]: SSL state (connect): before/connect initialization >> 2005.07.06 11:37:13 LOG7[17884:2]: SSL state (connect): SSLv3 write client hello A >> 2005.07.06 11:37:13 LOG7[17884:0]: Waiting 300 second(s) for 3 file descriptor(s) >> 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=4, (IN)->() >> 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=6, (IN)->() >> 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 2, FD=1, (IN)->(IN) >> >> The mail never gets delivered to either Thunderbird or outlook express. >> I get a certificate approval request from Thunderbird to which I grant, then nothing. If I disable ssl in thunderbird the mail gets accepted normally. >> >> What might I be doing wrong??? >> >> Thanks >> >> Doug P

1 0

Re: [stunnel-users] Stunnel for pop3 on solaris 2.6 ( Even more info)
by Douglas Phillipson 06 Jul '05

06 Jul '05

If I set "client = no" in stunnels config file I get the following from: openssl s_client -connect 172.20.12.59:995 CONNECTED(00000003) depth=0 /C=US/ST=Nevada/L=Las Vegas/O=Bechtel/OU=RSL/CN=test1 verify error:num=18:self signed certificate verify return:1 depth=0 /C=US/ST=Nevada/L=Las Vegas/O=Bechtel/OU=RSL/CN=test1 verify return:1 30463:error:140790E5:SSL routines:SSL23_WRITE:ssl handshake failure:s23_lib.c:226: Is a "self signed" Cert OK? I didn't sign anything though. Compiling stunnel created a cert. Installing openssl I think created a cert. Do these need to match somehow? If so how do you do that? I think I'm lost here... Regards Doug P Douglas Phillipson wrote: > I have a Solaris 2.6 box and am trying to get pop3 over SSL running. I use: > > qpopper 4.0.5 > openssl-0.9.7g > stunnel 4.10 compiled with gcc 2.95.3 > > When I compiled stunnel it made a private ket and certificate in /usr/local/etc/stunnel/stunnel.pem. > > Do I need anything else? > > I have the following configuration: > > Inetd.conf: > > pop3 stream tcp nowait root /usr/local/sbin/popper qpopper -S -t /poplog > > stunnel.conf: > > cert = /usr/local/etc/stunnel/stunnel.pem > key = /usr/local/etc/stunnel/stunnel.pem > debug = 7 > output = /stunnel.log > pid = /stunnel.pid > client = yes > > [pop3s] > accept = 995 > connect = 110 > > I run stunnel and get the following output: > > 2005.07.06 11:34:17 LOG5[17873:1]: stunnel 4.10 on sparc-sun-solaris2.6 UCONTEXT+POLL+IPv4 with OpenSSL 0.9.7g 11 Apr 2005 > 2005.07.06 11:34:18 LOG7[17873:1]: Snagged 64 random bytes from //.rnd > 2005.07.06 11:34:18 LOG7[17873:1]: Wrote 1024 new random bytes to //.rnd > 2005.07.06 11:34:18 LOG7[17873:1]: RAND_status claims sufficient entropy for the PRNG > 2005.07.06 11:34:18 LOG6[17873:1]: PRNG seeded successfully > 2005.07.06 11:34:18 LOG7[17873:1]: Certificate: /usr/local/etc/stunnel/stunnel.pem > 2005.07.06 11:34:18 LOG7[17873:1]: Key file: /usr/local/etc/stunnel/stunnel.pem > 2005.07.06 11:34:18 LOG6[17873:1]: file ulimit = 64 (can be changed with 'ulimit -n') > 2005.07.06 11:34:18 LOG6[17873:1]: poll() used - no FD_SETSIZE limit for file descriptors > 2005.07.06 11:34:18 LOG5[17873:1]: 29 clients allowed > 2005.07.06 11:34:18 LOG7[17873:1]: FD 4 in non-blocking mode > 2005.07.06 11:34:18 LOG7[17873:1]: FD 5 in non-blocking mode > 2005.07.06 11:34:18 LOG7[17873:1]: FD 6 in non-blocking mode > 2005.07.06 11:34:18 LOG7[17873:1]: SO_REUSEADDR option set on accept socket > 2005.07.06 11:34:18 LOG7[17873:1]: pop3s bound to 0.0.0.0:995 > 2005.07.06 11:34:18 LOG7[17874:1]: Created pid file /stunnel.pid > 2005.07.06 11:34:18 LOG7[17874:0]: Waiting -1 second(s) for 2 file descriptor(s) > > > I connect via pop3 in thunderbird with ssl and qpopper always says: > (null) at localhost (127.0.0.1): -ERR Unknown command: "". > (nulI/O error flushing output to client at localhost [127.0.0.1]: Broken pipe (32)l) at localhost (127.0.0.1): -ERR POP EOF or I/O Error > > Stunnel says: > > 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=4, (IN)->() > 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=6, (IN)->(IN) > 2005.07.06 11:37:13 LOG7[17884:1]: pop3s accepted FD=0 from 172.20.10.7:45464 > 2005.07.06 11:37:13 LOG7[17884:1]: Creating a new context > 2005.07.06 11:37:13 LOG7[17884:1]: Context 2 created > 2005.07.06 11:37:13 LOG7[17884:2]: pop3s started > 2005.07.06 11:37:13 LOG7[17884:2]: FD 0 in non-blocking mode > 2005.07.06 11:37:13 LOG7[17884:2]: TCP_NODELAY option set on local socket > 2005.07.06 11:37:13 LOG5[17884:2]: pop3s connected from 172.20.10.7:45464 > 2005.07.06 11:37:13 LOG7[17884:2]: FD 1 in non-blocking mode > 2005.07.06 11:37:13 LOG7[17884:2]: pop3s connecting 127.0.0.1:110 > 2005.07.06 11:37:13 LOG7[17884:2]: Remote FD=1 initialized > 2005.07.06 11:37:13 LOG7[17884:2]: TCP_NODELAY option set on remote socket > 2005.07.06 11:37:13 LOG7[17884:2]: SSL state (connect): before/connect initialization > 2005.07.06 11:37:13 LOG7[17884:2]: SSL state (connect): SSLv3 write client hello A > 2005.07.06 11:37:13 LOG7[17884:0]: Waiting 300 second(s) for 3 file descriptor(s) > 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=4, (IN)->() > 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=6, (IN)->() > 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 2, FD=1, (IN)->(IN) > > The mail never gets delivered to either Thunderbird or outlook express. > I get a certificate approval request from Thunderbird to which I grant, then nothing. If I disable ssl in thunderbird the mail gets accepted normally. > > What might I be doing wrong??? > > Thanks > > Doug P > _______________________________________________ > stunnel-users mailing list > stunnel-users(a)mirt.net > http://stunnel.mirt.net/mailman/listinfo/stunnel-users

1 0

Stunnel for pop3 on solaris 2.6
by Douglas Phillipson 06 Jul '05

06 Jul '05

I have a Solaris 2.6 box and am trying to get pop3 over SSL running. I use: qpopper 4.0.5 openssl-0.9.7g stunnel 4.10 compiled with gcc 2.95.3 When I compiled stunnel it made a private ket and certificate in /usr/local/etc/stunnel/stunnel.pem. Do I need anything else? I have the following configuration: Inetd.conf: pop3 stream tcp nowait root /usr/local/sbin/popper qpopper -S -t /poplog stunnel.conf: cert = /usr/local/etc/stunnel/stunnel.pem key = /usr/local/etc/stunnel/stunnel.pem debug = 7 output = /stunnel.log pid = /stunnel.pid client = yes [pop3s] accept = 995 connect = 110 I run stunnel and get the following output: 2005.07.06 11:34:17 LOG5[17873:1]: stunnel 4.10 on sparc-sun-solaris2.6 UCONTEXT+POLL+IPv4 with OpenSSL 0.9.7g 11 Apr 2005 2005.07.06 11:34:18 LOG7[17873:1]: Snagged 64 random bytes from //.rnd 2005.07.06 11:34:18 LOG7[17873:1]: Wrote 1024 new random bytes to //.rnd 2005.07.06 11:34:18 LOG7[17873:1]: RAND_status claims sufficient entropy for the PRNG 2005.07.06 11:34:18 LOG6[17873:1]: PRNG seeded successfully 2005.07.06 11:34:18 LOG7[17873:1]: Certificate: /usr/local/etc/stunnel/stunnel.pem 2005.07.06 11:34:18 LOG7[17873:1]: Key file: /usr/local/etc/stunnel/stunnel.pem 2005.07.06 11:34:18 LOG6[17873:1]: file ulimit = 64 (can be changed with 'ulimit -n') 2005.07.06 11:34:18 LOG6[17873:1]: poll() used - no FD_SETSIZE limit for file descriptors 2005.07.06 11:34:18 LOG5[17873:1]: 29 clients allowed 2005.07.06 11:34:18 LOG7[17873:1]: FD 4 in non-blocking mode 2005.07.06 11:34:18 LOG7[17873:1]: FD 5 in non-blocking mode 2005.07.06 11:34:18 LOG7[17873:1]: FD 6 in non-blocking mode 2005.07.06 11:34:18 LOG7[17873:1]: SO_REUSEADDR option set on accept socket 2005.07.06 11:34:18 LOG7[17873:1]: pop3s bound to 0.0.0.0:995 2005.07.06 11:34:18 LOG7[17874:1]: Created pid file /stunnel.pid 2005.07.06 11:34:18 LOG7[17874:0]: Waiting -1 second(s) for 2 file descriptor(s) I connect via pop3 in thunderbird with ssl and qpopper always says: (null) at localhost (127.0.0.1): -ERR Unknown command: "". (nulI/O error flushing output to client at localhost [127.0.0.1]: Broken pipe (32)l) at localhost (127.0.0.1): -ERR POP EOF or I/O Error Stunnel says: 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=4, (IN)->() 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=6, (IN)->(IN) 2005.07.06 11:37:13 LOG7[17884:1]: pop3s accepted FD=0 from 172.20.10.7:45464 2005.07.06 11:37:13 LOG7[17884:1]: Creating a new context 2005.07.06 11:37:13 LOG7[17884:1]: Context 2 created 2005.07.06 11:37:13 LOG7[17884:2]: pop3s started 2005.07.06 11:37:13 LOG7[17884:2]: FD 0 in non-blocking mode 2005.07.06 11:37:13 LOG7[17884:2]: TCP_NODELAY option set on local socket 2005.07.06 11:37:13 LOG5[17884:2]: pop3s connected from 172.20.10.7:45464 2005.07.06 11:37:13 LOG7[17884:2]: FD 1 in non-blocking mode 2005.07.06 11:37:13 LOG7[17884:2]: pop3s connecting 127.0.0.1:110 2005.07.06 11:37:13 LOG7[17884:2]: Remote FD=1 initialized 2005.07.06 11:37:13 LOG7[17884:2]: TCP_NODELAY option set on remote socket 2005.07.06 11:37:13 LOG7[17884:2]: SSL state (connect): before/connect initialization 2005.07.06 11:37:13 LOG7[17884:2]: SSL state (connect): SSLv3 write client hello A 2005.07.06 11:37:13 LOG7[17884:0]: Waiting 300 second(s) for 3 file descriptor(s) 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=4, (IN)->() 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 1, FD=6, (IN)->() 2005.07.06 11:37:13 LOG7[17884:0]: CONTEXT 2, FD=1, (IN)->(IN) The mail never gets delivered to either Thunderbird or outlook express. I get a certificate approval request from Thunderbird to which I grant, then nothing. If I disable ssl in thunderbird the mail gets accepted normally. What might I be doing wrong??? Thanks Doug P

1 4

Certificate renaming
by Paul Jones 05 Jul '05

05 Jul '05

So what you are saying is that Stunnel will still work successfully and find the correct certificates if I do NOT name the certificate files using their hash values? And since my server will only have a maximum of 16 clients, finding the correct certificate shouldn't cause too many speed issues. Right? Cheers. Paul _________________________________________________________________ Sell your car for $9 on carpoint.com.au http://www.carpoint.com.au/sellyourcar

1 0

Certificate renaming
by Paul Jones 04 Jul '05

04 Jul '05

Hi all, I understand that in order for Stunnel to find a certificate, it has to be renamed to something like xxxxxx.0 (where xxxxxx is obtained by using the "openssl x509" command on the certificate). Why is this the case? I read somewhere that this helps Stunnel find the certificate quicker (i.e. Stunnel does not need to look at every certificate file before it knows which one it wants). Can someone provide me a better explanation please? Is there any way around this? Renaming the certificate file in such a manner causes confusion when searching for the certificate of a particular client... Cheers. Paul _________________________________________________________________ REALESTATE: biggest buy/rent/share listings http://ninemsn.realestate.com.au

4 4

AW: [stunnel-users] Reporting problems
by Smith, Derek John 01 Jul '05

01 Jul '05

Hi all, sorry to have to use the list for this, but it would appear to be the last option available. This is from a previous posting from Brian Hatch ... > As the maintenance man of the 3.x branch (yes, some folks do > use it still) > the official stance is this: > > * if it's a security bug, I'll patch it. > * if it's a bug and you have a patch that works, I'll probably > apply it > * if it's a bug and you don't have a patch, I probably don't > have the cycles to investigate it, sorry. > However, the reality is different. Yes, some people still do use release 3. I have customers who prefer proven stability over new innovation. This is meant to be understood in a general way (i.e. database upgrades, OS updates etc) and not specifically to stunnel. I have mailed Brian several times about problems concerning release 3.26 and not a single reply has been received. Maybe the mails never arrived, but I doubt it very much. Also, no changes have been made to the software, so the supplied patch never got applied. In the complete absence of any proof to the contrary, I can only assume that the following statement holds:- RELEASE 3 IS DEAD !!!!!!!!!! I welcome evidence refuting this, but won't hold my breath. In the meantime, testing will start on release 4. Brian, if you're reading this, I'm not trying to start a flame war, or dissing you. A simple mail would have sufficed ... Derek

2 1

4.10 Crashes with Multiple Connections
by Eddie Atherton 30 Jun '05

30 Jun '05

Hi. Just started to use stunnel for NNTP traffic to Giga on a gateway Linux 2.4 box that sits between Windows and my cable modem. The NewsReader I use is configured to open 4 connections at the same time. Downloaded and installed 4.10, then immediately hit the -lpthread problem. Edited the Makefile, did a "make clean", "make", and "make install". Tried again. Got an immediate crash. Tried a few more times. Sometimes it crashed, other times the NewsReader reported one connection as working, one as connected, but not pulling any data, and two connections as connecting. So, it looks like everything works with only one active connection, but it segfaults when 2 or more threads are pulling data. The problem is, I can't always get it to crash. Plus, cpu use of stunnel sometimes pegs at 99%, even after closing the connections, and again this isn't consistent. When I turn debug on, it has always only used one active connection, so I don't get the crash. Plus, when debug is on, and I terminate the connections from the NewsReader, the log file expanded to 18meg in less than a minute. I then downloaded and installed 4.9, and this appears to handle the 4 active connections without any problems. Any suggestions as to what kind of information I should try and collect for this. Cheers, Eddie

1 0

SSL_read returned WANT_ - retry
by Leach, Rich 30 Jun '05

30 Jun '05

Greetings all- I'm an internet developer (mainly ColdFusion, some Windows, database stuff, etc.) and I've just inherited a Windows system that is utilizing Stunnel (I believe it's version 3.4) and in the log files I'm seeing this error all over the place: "SSL_read returned WANT_ - retry" It's usually listed around other messages in the log file such as "socket closed", "connection closed", etc. I noticed these messages in our logs because I got a call to restart our web services; everything froze up and the site would not come up. I stopped and started the web services and everything came back up. For a long time everyone around here has been blaming our ColdFusion server (version 6.1) but it looks to me like the Stunnel is throwing a wrench and halting everything - does anyone know what these errors mean or could you direct me to a place on the stunnel site that could offer some insight? I tried looking around the FAQ section but didn't see my error listed. Any help or input you could provide would be most appreciated! Many thanks in advance, Rich

1 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

stunnel-users