- stunnel-users - stunnel.org

Re (2): Problems with stunnel and cox.net on Windows 7
by peter＠easthope.ca 18 Jan '18

18 Jan '18

From: "Phil Smith III" <phs3stuff(a)cox.net> Date: Thu, 18 Jan 2018 12:47:47 -0500 > while the output= wasn't accepted (it said it wasn't valid there), Does the file exist? If not, can stunnel create it? Can you read it with an editor. Can you write into it interactively? > changing 127.0.0.1 to localhost seems to have fixed it! What does "nslookup localhost" report? Regards, ... P. -- 123456789 123456789 123456789 123456789 123456789 123456789 123456789 Tel: +1 360 639 0202 Pender Is.: +1 250 629 3757 http://easthope.ca/Peter.html Bcc: peter at easthope. ca

2 1

Re (2): Problems with stunnel and cox.net on Windows 7
by peter＠easthope.ca 18 Jan '18

18 Jan '18

From: "Phil Smith III" <phs3stuff(a)cox.net> Date: Thu, 18 Jan 2018 12:47:47 -0500 > while the output= wasn't accepted (it said it wasn't valid there), Does the file exist? If not, can stunnel create it? Can you read it with an editor. Can you write into it interactively? What does "ls -ld C:\temp\stunnel-log.txt" or Win equivalent give? > changing 127.0.0.1 to localhost seems to have fixed it! What does "nslookup localhost" report? Regards, ... P. -- 123456789 123456789 123456789 123456789 123456789 123456789 123456789 Tel: +1 360 639 0202 Pender Is.: +1 250 629 3757 http://easthope.ca/Peter.html Bcc: peter at easthope. ca

1 0

Re: [stunnel-users] Problems with stunnel and cox.net on Windows 7
by Phil Smith III 18 Jan '18

18 Jan '18

Well, I'm stunned (pun intended): while the output= wasn't accepted (it said it wasn't valid there), changing 127.0.0.1 to localhost seems to have fixed it! -----Original Message----- From: stunnel-users [mailto:stunnel-users-bounces@stunnel.org] On Behalf Of peter(a)easthope.ca Sent: Thursday, January 18, 2018 12:20 AM To: stunnel-users(a)stunnel.org Subject: Re: [stunnel-users] Problems with stunnel and cox.net on Windows 7 From: "Phil Smith III" <phs3stuff(a)cox.net> Date: Wed, 17 Jan 2018 21:59:42 -0500 > Here's my stunnel.conf, with blank lines and comments removed: > client = yes > output = C:\temp\stunnel-log.txt > taskbar = yes > debug = 7 > [POP3 Incoming] > protocol = pop3 > accept = 127.0.0.1:1110 > connect = pop.cox.net:995 > protocol = smtp > accept = 127.0.0.1:25 > connect = smtp.cox.net:465 Here it worked for POP3 immediately. Not needed for SMTP. https://en.wikibooks.org/wiki/Oberon/Oberon#Receiving_and_Sending_Email Before using 1110, check. netstat -tulpn | grep 1110 Correct id, passwd and port are in the MUA? Separate configurations for POP3 and SMTP. Troubleshoot one at a time. "protocol" shouldn't be necessary. Try this. [POP3] client = yes output = C:\temp\stunnel-log.txt accept = localhost:1110 connect = pop.cox.net:995 Regards, ... Peter E. -- 123456789 123456789 123456789 123456789 123456789 123456789 123456789 Tel: +1 360 639 0202 Pender Is.: +1 250 629 3757 http://easthope.ca/Peter.html Bcc: peter at easthope. ca _______________________________________________ stunnel-users mailing list stunnel-users(a)stunnel.org https://www.stunnel.org/cgi-bin/mailman/listinfo/stunnel-users

1 0

Re (2): Problems with stunnel and cox.net on Windows 7
by peter＠easthope.ca 18 Jan '18

18 Jan '18

From: Andrew Culver <aculver(a)uwo.ca> Date: Thu, 18 Jan 2018 11:39:20 -0500 > Some protocols use a protocol-specific command to begin TLS negotiation > (STARTTLS for SMTP and POP3), so stunnel needs to know how (and be told) to > use that command. Phil, General troubleshooting rule, always try for a most elementary, ordinary case before resorting to a special case. Regards, ... Peter E. -- 123456789 123456789 123456789 123456789 123456789 123456789 123456789 Tel: +1 360 639 0202 Pender Is.: +1 250 629 3757 http://easthope.ca/Peter.html Bcc: peter at easthope. ca

1 0

Re: [stunnel-users] Problems with stunnel and cox.net on Windows 7
by peter＠easthope.ca 18 Jan '18

18 Jan '18

P.s. From: "Phil Smith III" <phs3stuff(a)cox.net> Date: Wed, 17 Jan 2018 21:59:42 -0500 > protocol = pop3 The stunnel(8) manual has this. "protocol = PROTO application protocol to negotiate TLS This option enables initial, protocol-specific negotiation of the TLS encryption." Is there any need for protocol-specific negotiation? Why? Stunnel provides a tunnel and POP3 works inside. That's all. Try the configuration without the protocol option. Regards, ... Peter E. -- 123456789 123456789 123456789 123456789 123456789 123456789 123456789 Tel: +1 360 639 0202 Pender Is.: +1 250 629 3757 http://easthope.ca/Peter.html Bcc: peter at easthope. ca

2 1

Re: [stunnel-users] Problems with stunnel and cox.net on Windows 7
by Phil Smith III 18 Jan '18

18 Jan '18

Right, I had done that: >Yes, I have protocol = pop3 and protocol = smtp; I’ve tried >commenting them out individually, no change that I can see. The doc really isn't clear about why you'd want/need these, by the way. But your earlier suggestion (localhost instead of 127.0.0.1) seems to have worked. Still a mystery. I'll be tinkering on machine2 tonight, and then will try on the wife's machine! -----Original Message----- From: stunnel-users [mailto:stunnel-users-bounces@stunnel.org] On Behalf Of peter(a)easthope.ca Sent: Thursday, January 18, 2018 9:51 AM To: stunnel-users(a)stunnel.org Subject: Re: [stunnel-users] Problems with stunnel and cox.net on Windows 7 P.s. From: "Phil Smith III" <phs3stuff(a)cox.net> Date: Wed, 17 Jan 2018 21:59:42 -0500 > protocol = pop3 The stunnel(8) manual has this. "protocol = PROTO application protocol to negotiate TLS This option enables initial, protocol-specific negotiation of the TLS encryption." Is there any need for protocol-specific negotiation? Why? Stunnel provides a tunnel and POP3 works inside. That's all. Try the configuration without the protocol option. Regards, ... Peter E. -- 123456789 123456789 123456789 123456789 123456789 123456789 123456789 Tel: +1 360 639 0202 Pender Is.: +1 250 629 3757 http://easthope.ca/Peter.html Bcc: peter at easthope. ca _______________________________________________ stunnel-users mailing list stunnel-users(a)stunnel.org https://www.stunnel.org/cgi-bin/mailman/listinfo/stunnel-users

1 0

Re: [stunnel-users] Problems with stunnel and cox.net on Windows 7
by peter＠easthope.ca 18 Jan '18

18 Jan '18

From: "Phil Smith III" <phs3stuff(a)cox.net> Date: Wed, 17 Jan 2018 21:59:42 -0500 > Here's my stunnel.conf, with blank lines and comments removed: > client = yes > output = C:\temp\stunnel-log.txt > taskbar = yes > debug = 7 > [POP3 Incoming] > protocol = pop3 > accept = 127.0.0.1:1110 > connect = pop.cox.net:995 > protocol = smtp > accept = 127.0.0.1:25 > connect = smtp.cox.net:465 Here it worked for POP3 immediately. Not needed for SMTP. https://en.wikibooks.org/wiki/Oberon/Oberon#Receiving_and_Sending_Email Before using 1110, check. netstat -tulpn | grep 1110 Correct id, passwd and port are in the MUA? Separate configurations for POP3 and SMTP. Troubleshoot one at a time. "protocol" shouldn't be necessary. Try this. [POP3] client = yes output = C:\temp\stunnel-log.txt accept = localhost:1110 connect = pop.cox.net:995 Regards, ... Peter E. -- 123456789 123456789 123456789 123456789 123456789 123456789 123456789 Tel: +1 360 639 0202 Pender Is.: +1 250 629 3757 http://easthope.ca/Peter.html Bcc: peter at easthope. ca

1 0

Problems with stunnel and cox.net on Windows 7
by Phil Smith III 18 Jan '18

18 Jan '18

Cox has been warning us that as of next week, we need to use encrypted POP and SMTP. I have two reasons to use stunnel now: 1) I use the K9 Bayesian email proxy for spam filtering, and it doesn't do TLS 2) My wife still uses Eudora and doesn't want to switch, and while Eudora allegedly supports TLS, it's not trivial to make it work (and Cox certainly won't help) Thus stunnel looks like an ideal solution. But after tinkering for a few hours on two machines, I'm stumped. While I can type an SMTP or POP transaction in my sleep, I'm not so good at typing a TLS handshake, so I'm using Outlook 2013 as my test platform (and that's what I'll be using myself later anyway). I setup a new account, point to the right ports on 127.0.0.1, and the test times out, with: Log onto incoming mail server (POP3): The operation timed out waiting for a response from the receiving (POP) server. If you continue to receive this message, contact your server administrator or Internet service provider (ISP). Send test e-mail message: The operation timed out waiting for a response from the sending (SMTP) server. If you continue to receive this message, contact your server administrator or Internet service provider (ISP). Here's my stunnel.conf, with blank lines and comments removed: client = yes output = C:\temp\stunnel-log.txt taskbar = yes debug = 7 [POP3 Incoming] protocol = pop3 accept = 127.0.0.1:1110 connect = pop.cox.net:995 protocol = smtp accept = 127.0.0.1:25 connect = smtp.cox.net:465 Yes, that's port 1110, since K9 is using 110. Shouldn't matter, right? (As long as I point the client at 1110, obviously!) Yes, I have protocol = pop3 and protocol = smtp; I've tried commenting them out individually, no change that I can see. Yes, I'm running stunnel explicitly, not as a service-I found threads suggesting that it just doesn't work as a service on Windows 7 for some reason (and in fact beat my head against that wall for a while first). And here's the stunnel log: 2018.01.17 21:29:24 LOG7[main]: Running on Windows 6.1 2018.01.17 21:29:24 LOG7[main]: No limit detected for the number of clients 2018.01.17 21:29:24 LOG5[main]: stunnel 5.44 on x86-pc-msvc-1500 platform 2018.01.17 21:29:24 LOG5[main]: Compiled/running with OpenSSL 1.0.2m-fips 2 Nov 2017 2018.01.17 21:29:24 LOG5[main]: Threading:WIN32 Sockets:SELECT,IPv6 TLS:ENGINE,FIPS,OCSP,PSK,SNI 2018.01.17 21:29:24 LOG7[main]: errno: (*_errno()) 2018.01.17 21:29:24 LOG7[ui]: GUI message loop initialized 2018.01.17 21:29:24 LOG7[main]: Running on Windows 6.1 2018.01.17 21:29:24 LOG5[main]: Reading configuration from file stunnel.conf 2018.01.17 21:29:24 LOG5[main]: UTF-8 byte order mark not detected 2018.01.17 21:29:24 LOG5[main]: FIPS mode disabled 2018.01.17 21:29:24 LOG7[main]: Compression disabled 2018.01.17 21:29:24 LOG7[main]: Snagged 64 random bytes from C:/.rnd 2018.01.17 21:29:24 LOG7[main]: Wrote 0 new random bytes to C:/.rnd 2018.01.17 21:29:24 LOG7[main]: PRNG seeded successfully 2018.01.17 21:29:24 LOG6[main]: Initializing service [POP3 Incoming] 2018.01.17 21:29:24 LOG7[main]: Ciphers: HIGH:!DH:!aNULL:!SSLv2 2018.01.17 21:29:24 LOG7[main]: TLS options: 0x03000004 (+0x03000000, -0x00000000) 2018.01.17 21:29:24 LOG7[main]: No certificate or private key specified 2018.01.17 21:29:24 LOG4[main]: Service [POP3 Incoming] needs authentication to prevent MITM attacks 2018.01.17 21:29:24 LOG5[main]: Configuration successful 2018.01.17 21:29:24 LOG7[main]: Binding service [POP3 Incoming] 2018.01.17 21:29:24 LOG7[main]: Listening file descriptor created (FD=460) 2018.01.17 21:29:24 LOG7[main]: Option SO_EXCLUSIVEADDRUSE set on accept socket 2018.01.17 21:29:24 LOG7[main]: Service [POP3 Incoming] (FD=460) bound to 127.0.0.1:1110 2018.01.17 21:29:24 LOG7[main]: Listening file descriptor created (FD=464) 2018.01.17 21:29:24 LOG7[main]: Option SO_EXCLUSIVEADDRUSE set on accept socket 2018.01.17 21:29:24 LOG7[main]: Service [POP3 Incoming] (FD=464) bound to 127.0.0.1:25 2018.01.17 21:29:24 LOG7[cron]: Cron thread initialized 2018.01.17 21:29:50 LOG7[main]: Found 1 ready file descriptor(s) 2018.01.17 21:29:50 LOG7[main]: FD=424 ifds=r-x ofds=--- 2018.01.17 21:29:50 LOG7[main]: FD=460 ifds=r-x ofds=r-- 2018.01.17 21:29:50 LOG7[main]: Service [POP3 Incoming] accepted (FD=488) from 127.0.0.1:54855 2018.01.17 21:29:50 LOG7[main]: Creating a new thread 2018.01.17 21:29:50 LOG7[main]: New thread created 2018.01.17 21:29:50 LOG7[0]: Service [POP3 Incoming] started 2018.01.17 21:29:50 LOG7[0]: Option TCP_NODELAY set on local socket 2018.01.17 21:29:50 LOG5[0]: Service [POP3 Incoming] accepted connection from 127.0.0.1:54855 2018.01.17 21:29:50 LOG6[0]: failover: round-robin, starting at entry #1 2018.01.17 21:29:50 LOG6[0]: s_connect: connecting 68.6.19.8:465 2018.01.17 21:29:50 LOG7[0]: s_connect: s_poll_wait 68.6.19.8:465: waiting 10 seconds 2018.01.17 21:29:50 LOG5[0]: s_connect: connected 68.6.19.8:465 2018.01.17 21:29:50 LOG5[0]: Service [POP3 Incoming] connected remote server from 192.168.1.17:54856 2018.01.17 21:29:50 LOG7[0]: Option TCP_NODELAY set on remote socket 2018.01.17 21:29:50 LOG7[0]: Remote descriptor (FD=508) initialized 2018.01.17 21:30:24 LOG6[cron]: Executing cron jobs 2018.01.17 21:30:24 LOG6[cron]: Cron jobs completed in 0 seconds 2018.01.17 21:30:24 LOG7[cron]: Waiting 86400 seconds 2018.01.17 21:31:05 LOG7[main]: Found 1 ready file descriptor(s) 2018.01.17 21:31:05 LOG7[main]: FD=424 ifds=r-x ofds=--- 2018.01.17 21:31:05 LOG7[main]: FD=460 ifds=r-x ofds=--- 2018.01.17 21:31:05 LOG7[main]: Service [POP3 Incoming] accepted (FD=528) from 127.0.0.1:54891 2018.01.17 21:31:05 LOG7[main]: Creating a new thread 2018.01.17 21:31:05 LOG7[main]: New thread created 2018.01.17 21:31:05 LOG7[1]: Service [POP3 Incoming] started 2018.01.17 21:31:05 LOG7[1]: Option TCP_NODELAY set on local socket 2018.01.17 21:31:05 LOG5[1]: Service [POP3 Incoming] accepted connection from 127.0.0.1:54891 2018.01.17 21:31:05 LOG6[1]: failover: round-robin, starting at entry #0 2018.01.17 21:31:05 LOG6[1]: s_connect: connecting 146.20.147.245:995 2018.01.17 21:31:05 LOG7[1]: s_connect: s_poll_wait 146.20.147.245:995: waiting 10 seconds 2018.01.17 21:31:05 LOG5[1]: s_connect: connected 146.20.147.245:995 2018.01.17 21:31:05 LOG5[1]: Service [POP3 Incoming] connected remote server from 192.168.1.17:54892 2018.01.17 21:31:05 LOG7[1]: Option TCP_NODELAY set on remote socket 2018.01.17 21:31:05 LOG7[1]: Remote descriptor (FD=336) initialized 2018.01.17 21:34:05 LOG3[1]: Unexpected socket close (s_read) 2018.01.17 21:34:05 LOG5[1]: Connection reset: 0 byte(s) sent to TLS, 0 byte(s) sent to socket 2018.01.17 21:34:05 LOG7[1]: Remote descriptor (FD=336) closed 2018.01.17 21:34:05 LOG7[1]: Local descriptor (FD=528) closed 2018.01.17 21:34:05 LOG7[1]: Service [POP3 Incoming] finished (1 left) 2018.01.17 21:34:05 LOG7[1]: str_stats: 1 block(s), 32 data byte(s), 58 control byte(s) 2018.01.17 21:34:05 LOG7[1]: str_stats: 32 byte(s) at ..\src\network.c:680 2018.01.17 21:34:50 LOG6[0]: s_read: s_poll_wait: TIMEOUTbusy exceeded: sending reset 2018.01.17 21:34:50 LOG5[0]: Connection reset: 0 byte(s) sent to TLS, 0 byte(s) sent to socket 2018.01.17 21:34:50 LOG7[0]: Remote descriptor (FD=508) closed 2018.01.17 21:34:50 LOG7[0]: Local descriptor (FD=488) closed 2018.01.17 21:34:50 LOG7[0]: Service [POP3 Incoming] finished (0 left) 2018.01.17 21:34:50 LOG7[0]: str_stats: 1 block(s), 32 data byte(s), 58 control byte(s) 2018.01.17 21:34:50 LOG7[0]: str_stats: 32 byte(s) at ..\src\network.c:680 It looks like it connects, but then just sits there?! I see these: No certificate or private key specified Service [POP3 Incoming] needs authentication to prevent MITM attacks but that's during startup. Or are those significant? I feel like I'm one setting away from having this all work.! Thanks in advance for any suggestions. -- ...phsiii

1 0

TLS socket closed (SSL_write) with unsent byte(s)
by Cotswold Group - Stephen Murray 16 Jan '18

16 Jan '18

Hi All, We use stunnel to send emails from our scanners to our hosted mail provider (that now only accept TLS connections). This has been working well for a number of years, but just recently the scanners (all Lexmark) have sporadically been giving errors of 'connection failed'. If I check the stunnel logs, the failed send has an extra line in of TLS socket closed (SSL_write) with 418 unsent byte(s) Below are 2 examples, 59 was sent successfully, 60 was not. 2018.01.16 14:37:24 LOG5[59]: Service [fasthosts-smtp] accepted connection from 10.226.174.95:11473 2018.01.16 14:37:24 LOG5[59]: s_connect: connected 213.171.216.50:465 2018.01.16 14:37:24 LOG5[59]: Service [fasthosts-smtp] connected remote server from 10.226.150.9:52577 2018.01.16 14:37:24 LOG5[59]: OCSP: Connecting the AIA responder "http://t2.symcb.com" 2018.01.16 14:37:24 LOG5[59]: s_connect: connected 23.55.155.27:80 2018.01.16 14:37:24 LOG5[59]: OCSP: Certificate accepted 2018.01.16 14:37:24 LOG5[59]: OCSP: Connecting the AIA responder "http://tj.symcd.com" 2018.01.16 14:37:25 LOG5[59]: s_connect: connected 23.55.155.27:80 2018.01.16 14:37:25 LOG5[59]: OCSP: Certificate accepted 2018.01.16 14:37:25 LOG5[59]: Certificate accepted at depth=0: C=GB, ST=Gloucestershire, L=Gloucester, O=Fasthosts Internet Limited, OU=NOC, CN=smtp.livemail.co.uk 2018.01.16 14:37:29 LOG5[59]: Connection closed: 167106 byte(s) sent to TLS, 376 byte(s) sent to socket 2018.01.16 14:38:19 LOG5[60]: Service [fasthosts-smtp] accepted connection from 10.226.174.95:11483 2018.01.16 14:38:19 LOG5[60]: s_connect: connected 213.171.216.50:465 2018.01.16 14:38:19 LOG5[60]: Service [fasthosts-smtp] connected remote server from 10.226.150.9:52620 2018.01.16 14:38:19 LOG5[60]: OCSP: Connecting the AIA responder "http://t2.symcb.com" 2018.01.16 14:38:19 LOG5[60]: s_connect: connected 23.55.155.27:80 2018.01.16 14:38:19 LOG5[60]: OCSP: Certificate accepted 2018.01.16 14:38:19 LOG5[60]: OCSP: Connecting the AIA responder "http://tj.symcd.com" 2018.01.16 14:38:19 LOG5[60]: s_connect: connected 23.55.155.27:80 2018.01.16 14:38:19 LOG5[60]: OCSP: Certificate accepted 2018.01.16 14:38:19 LOG5[60]: Certificate accepted at depth=0: C=GB, ST=Gloucestershire, L=Gloucester, O=Fasthosts Internet Limited, OU=NOC, CN=smtp.livemail.co.uk 2018.01.16 14:38:24 LOG3[60]: TLS socket closed (SSL_write) with 418 unsent byte(s) 2018.01.16 14:38:24 LOG5[60]: Connection reset: 964 byte(s) sent to TLS, 392 byte(s) sent to socket Has anyone come across this before? Our mail provider has send limits in place of 50 every 10 minutes, but I am confident we are not reaching those amounts. Any help appreciated. Regards, Stephen

1 0

Connecting a HTTP browser to a HTTPS server.
by peter＠easthope.ca 12 Jan '18

12 Jan '18

Hello, I'm interested to configure stunnel to allow my local non-TLS Web browser to receive pages from a remote server providing only HTTPS. This section of https://www.stunnel.org/static/stunnel.html appears relevant. transparent = none | source | destination | both (Unix only) enable transparent proxy support on selected platforms ... destination The original destination is used instead of the connect option. A service section for transparent destination may look like this: [transparent] client = yes accept = <stunnel_port> transparent = destination This configuration requires iptables setup to work, possibly in /etc/rc.local or equivalent file. ... For a connect target installed on a remote host: /sbin/iptables -I INPUT -i eth0 -p tcp --dport <stunnel_port> -j ACCEPT /sbin/iptables -t nat -I PREROUTING -p tcp --dport <redirected_port> \ -i eth0 -j DNAT --to-destination <local_ip>:<stunnel_port> The transparent destination option is currently only supported on Linux. ==================================== Substitution of parameters for my context gives these configurations.. [transparent] client = yes accept = 127.0.0.1:954 transparent = destination /sbin/iptables -I INPUT -i eth0 -p tcp --dport 954 -j ACCEPT /sbin/iptables -t nat -I PREROUTING -p tcp --dport 443 \ -i eth0 -j DNAT --to-destination 127.0.0.1:954 The non-TLS browser would address http://www.stunnel.org:954/ and receive the page from https://www.stunnel.org:443/. According to Wikipedia, 954 is not assigned. The browser would continue to receive the page http://www.website.org/ without involvement of stunnel or iptables. Plausible? Thanks, ... P. -- 123456789 123456789 123456789 123456789 123456789 123456789 123456789 Tel: +1 360 639 0202 Pender Is.: +1 250 629 3757 http://easthope.ca/Peter.html Bcc: peter at easthope. ca

1 0

Versions
by Editor 04 Jan '18

04 Jan '18

Hi: On the various repo sites for Redhat-CentOS-Oracle Linux, etc., they are only providing version 4.29 of stunnel. Is there anyway to push them an updated version? I normally use these repo providers: Oracle Linux repo EPEL 6 Repo rpmforce Repo In the end, I tried to compile the latest version on Linux but I get this somewhat cryptic message about a TLS installation directory. I never had this issue before but to be honest, I normally used Solaris (8-9-10) in the past. "Could not find your TLS library installation dir Use --with-ssl option to fix this problem" I tried pointing it at /usr/bin/openssl and also at /usr/lib64/openssl without any luck. Sorry for bothering the list if this has been answered in the past. Thanks. Kevin

2 1

Re: [stunnel-users] outgoing mail solution?
by Parker Brooks 02 Jan '18

02 Jan '18

Douglas, it worked! Thank you! I had not turned on gmail's allow less secure apps, so my first try netted this 2018.01.02 09:00:27 LOG5[main]: stunnel 5.44 on x86-pc-msvc-1500 platform 2018.01.02 09:00:27 LOG5[main]: Compiled/running with OpenSSL 1.0.2m-fips 2 Nov 2017 2018.01.02 09:00:27 LOG5[main]: Threading:WIN32 Sockets:SELECT,IPv6 TLS:ENGINE,FIPS,OCSP,PSK,SNI 2018.01.02 09:00:27 LOG5[main]: Reading configuration from file stunnel.conf 2018.01.02 09:00:27 LOG5[main]: UTF-8 byte order mark not detected 2018.01.02 09:00:27 LOG5[main]: FIPS mode disabled 2018.01.02 09:00:27 LOG5[main]: Configuration successful 2018.01.02 09:03:09 LOG5[0]: Service [gmail-smtp] accepted connection from 127.0.0.1:1046 2018.01.02 09:03:09 LOG5[0]: s_connect: connected 173.194.78.108:587 2018.01.02 09:03:09 LOG5[0]: Service [gmail-smtp] connected remote server from 192.168.1.103:1047 2018.01.02 09:03:09 LOG5[0]: OCSP: Connecting the AIA responder " http://ocsp.pki.goog/gsr2" 2018.01.02 09:03:10 LOG5[0]: s_connect: connected 172.217.12.78:80 2018.01.02 09:03:10 LOG5[0]: OCSP: Certificate accepted 2018.01.02 09:03:10 LOG5[0]: OCSP: Connecting the AIA responder " http://ocsp.pki.goog/GTSGIAG3" 2018.01.02 09:03:10 LOG5[0]: s_connect: connected 172.217.12.78:80 2018.01.02 09:03:10 LOG5[0]: OCSP: Certificate accepted 2018.01.02 09:03:10 LOG5[0]: Certificate accepted at depth=0: C=US, ST=California, L=Mountain View, O=Google Inc, CN=smtp.gmail.com 2018.01.02 09:03:10 LOG3[0]: PLAIN Authentication Failed 2018.01.02 09:03:10 LOG5[0]: Connection reset: 0 byte(s) sent to TLS, 0 byte(s) sent to socket my monitoring s/w showed To = xxxxxx(a)gmail.com Connecting to mail server... 127.0.0.1 E-mail Failed. Error: 25058 / [10057] Socket is not connected 1/2/2018 9:03:10 AM there was a security alert from gmail with the instructions to enabling the less secure apps. turned it on and tried again... 2018.01.02 09:17:45 LOG5[1]: Service [gmail-smtp] accepted connection from 127.0.0.1:1050 2018.01.02 09:17:45 LOG5[1]: s_connect: connected 173.194.78.109:587 2018.01.02 09:17:45 LOG5[1]: Service [gmail-smtp] connected remote server from 192.168.1.103:1051 2018.01.02 09:17:47 LOG5[1]: Connection closed: 306 byte(s) sent to TLS, 508 byte(s) sent to socket again, thank you! On Tue, Jan 2, 2018 at 7:35 AM, Harris, Douglas (SSC/SPC) < douglas.harris(a)canada.ca> wrote: > Try this… > > > > [gmail-smtp] > > client = yes > > accept = 127.0.0.1:25 > > connect = smtp.gmail.com:587 <http://smtp.gmail.com:465> > > verifyChain = yes > > CAfile = ca-certs.pem > > checkHost = smtp.gmail.com > > OCSPaia = yes > > protocol = smtp > > protocolUsername = [your gmail userid] > > protocolPassword = [your gmail password] > > > > > > > > > > *From:* stunnel-users [mailto:stunnel-users-bounces@stunnel.org] *On > Behalf Of *Parker Brooks > *Sent:* January 2, 2018 12:34 AM > *To:* stunnel-users(a)stunnel.org > *Subject:* [stunnel-users] outgoing mail solution? > > > > hello and happy new year! > > > > it seemed this question has been asked before in various ways, however > with the frequent version updates i'm still unclear if stunnel is meant as > a fix to my issue. > > > > i have a outdated home monitoring program running on a winxp box. when an > error occurs it sends an email. it's setup screen (besides the destination > email field) only have a) smtp server, b) undeliverable email address and > c) port # fields to configure. > > > > my workaround for several years (when port 25 became blocked or needed > authetication) has been to use LOApost but they've recently ceased. that > program, also running in the background, and if i understand correctly > acted as a local host. i configured the monitoring program's smtp for > 127.0.0.1 and LOApost did the rest > > > > other searches suggest cloud solutions like sendgrid and authsmtp. but > those sites need a login/ password with the outgoing message. perhaps this > is where i'm getting confused between the differences of SSL and smtp > authentication? > > > > i have gotten gsuite's restricted smtp server to work, but i hoped stunnel > would provide a better solution (ability to send besides gmail/ gsuite > addresses). > > > > below are my conf and log files of my attempt. again, i can't tell if i've > missed a step, or if this is not intended to work. > > > > PS: i've also tried to understand if this feature is on the todo list? > > tia, > > parker > > > > stunnel.conf > > > > [gmail-pop3] > > client = yes > > accept = 127.0.0.1:110 > > connect = pop.gmail.com:995 > > verifyChain = yes > > CAfile = ca-certs.pem > > checkHost = pop.gmail.com > > OCSPaia = yes > > > > [gmail-imap] > > client = yes > > accept = 127.0.0.1:143 > > connect = imap.gmail.com:993 > > verifyChain = yes > > CAfile = ca-certs.pem > > checkHost = imap.gmail.com > > OCSPaia = yes > > > > [gmail-smtp] > > client = yes > > accept = 127.0.0.1:25 > > connect = smtp.gmail.com:465 > > verifyChain = yes > > CAfile = ca-certs.pem > > checkHost = smtp.gmail.com > > OCSPaia = yes > > > > log file > > > > 2017.12.31 16:22:45 LOG5[main]: stunnel 5.44 on x86-pc-msvc-1500 platform > > 2017.12.31 16:22:45 LOG5[main]: Compiled/running with OpenSSL 1.0.2m-fips > 2 Nov 2017 > > 2017.12.31 16:22:45 LOG5[main]: Threading:WIN32 Sockets:SELECT,IPv6 > TLS:ENGINE,FIPS,OCSP,PSK,SNI > > 2017.12.31 16:22:45 LOG5[main]: Reading configuration from file > stunnel.conf > > 2017.12.31 16:22:45 LOG5[main]: UTF-8 byte order mark not detected > > 2017.12.31 16:22:45 LOG5[main]: FIPS mode disabled > > 2017.12.31 16:22:45 LOG5[main]: Configuration successful > > 2017.12.31 16:25:03 LOG5[0]: Service [gmail-smtp] accepted connection from > 127.0.0.1:1141 > > 2017.12.31 16:25:03 LOG5[0]: s_connect: connected 173.194.67.108:465 > > 2017.12.31 16:25:03 LOG5[0]: Service [gmail-smtp] connected remote server > from 192.168.1.103:1142 > > 2017.12.31 16:25:03 LOG5[0]: OCSP: Connecting the AIA responder " > http://ocsp.pki.goog/gsr2" > > 2017.12.31 16:25:04 LOG5[0]: s_connect: connected 172.217.14.174:80 > > 2017.12.31 16:25:04 LOG5[0]: OCSP: Certificate accepted > > 2017.12.31 16:25:04 LOG5[0]: OCSP: Connecting the AIA responder " > http://ocsp.pki.goog/GTSGIAG3" > > 2017.12.31 16:25:04 LOG5[0]: s_connect: connected 172.217.14.174:80 > > 2017.12.31 16:25:04 LOG5[0]: OCSP: Certificate accepted > > 2017.12.31 16:25:04 LOG5[0]: Certificate accepted at depth=0: C=US, > ST=California, L=Mountain View, O=Google Inc, CN=smtp.gmail.com > > 2017.12.31 16:25:04 LOG5[0]: Connection closed: 55 byte(s) sent to TLS, > 514 byte(s) sent to socket > > > > > > and this is what shows up on my monitoring program log file... > > > > To = xxxxxx(a)gmail.com > > Connecting to mail server... 127.0.0.1 > > Sending message... > > E-mail Failed. Error: 20162 / 530 5.5.1 https://support.google.com/ > mail/?p=WantAuthError 51sm9558225oti.10 - gsmtp > > 12/31/2017 4:25:04 PM >

1 0

Re: [stunnel-users] outgoing mail solution?
by Josealf.rm 02 Jan '18

02 Jan '18

Cool! I was not aware that stunnel fully supported SMTP authentication. Thanks Parker! Saludos Jose Alfredo Diaz > On Jan 2, 2018, at 8:35 AM, Harris, Douglas (SSC/SPC) <douglas.harris(a)canada.ca> wrote: > > Try this… > > [gmail-smtp] > client = yes > accept = 127.0.0.1:25 > connect = smtp.gmail.com:587 > verifyChain = yes > CAfile = ca-certs.pem > checkHost = smtp.gmail.com > OCSPaia = yes > protocol = smtp > protocolUsername = [your gmail userid] > protocolPassword = [your gmail password] > > > > > From: stunnel-users [mailto:stunnel-users-bounces@stunnel.org] On Behalf Of Parker Brooks > Sent: January 2, 2018 12:34 AM > To: stunnel-users(a)stunnel.org > Subject: [stunnel-users] outgoing mail solution? > > hello and happy new year! > > it seemed this question has been asked before in various ways, however with the frequent version updates i'm still unclear if stunnel is meant as a fix to my issue. > > i have a outdated home monitoring program running on a winxp box. when an error occurs it sends an email. it's setup screen (besides the destination email field) only have a) smtp server, b) undeliverable email address and c) port # fields to configure. > > my workaround for several years (when port 25 became blocked or needed authetication) has been to use LOApost but they've recently ceased. that program, also running in the background, and if i understand correctly acted as a local host. i configured the monitoring program's smtp for 127.0.0.1 and LOApost did the rest > > other searches suggest cloud solutions like sendgrid and authsmtp. but those sites need a login/ password with the outgoing message. perhaps this is where i'm getting confused between the differences of SSL and smtp authentication? > > i have gotten gsuite's restricted smtp server to work, but i hoped stunnel would provide a better solution (ability to send besides gmail/ gsuite addresses). > > below are my conf and log files of my attempt. again, i can't tell if i've missed a step, or if this is not intended to work. > > PS: i've also tried to understand if this feature is on the todo list? > tia, > parker > > stunnel.conf > > [gmail-pop3] > client = yes > accept = 127.0.0.1:110 > connect = pop.gmail.com:995 > verifyChain = yes > CAfile = ca-certs.pem > checkHost = pop.gmail.com > OCSPaia = yes > > [gmail-imap] > client = yes > accept = 127.0.0.1:143 > connect = imap.gmail.com:993 > verifyChain = yes > CAfile = ca-certs.pem > checkHost = imap.gmail.com > OCSPaia = yes > > [gmail-smtp] > client = yes > accept = 127.0.0.1:25 > connect = smtp.gmail.com:465 > verifyChain = yes > CAfile = ca-certs.pem > checkHost = smtp.gmail.com > OCSPaia = yes > > log file > > 2017.12.31 16:22:45 LOG5[main]: stunnel 5.44 on x86-pc-msvc-1500 platform > 2017.12.31 16:22:45 LOG5[main]: Compiled/running with OpenSSL 1.0.2m-fips 2 Nov 2017 > 2017.12.31 16:22:45 LOG5[main]: Threading:WIN32 Sockets:SELECT,IPv6 TLS:ENGINE,FIPS,OCSP,PSK,SNI > 2017.12.31 16:22:45 LOG5[main]: Reading configuration from file stunnel.conf > 2017.12.31 16:22:45 LOG5[main]: UTF-8 byte order mark not detected > 2017.12.31 16:22:45 LOG5[main]: FIPS mode disabled > 2017.12.31 16:22:45 LOG5[main]: Configuration successful > 2017.12.31 16:25:03 LOG5[0]: Service [gmail-smtp] accepted connection from 127.0.0.1:1141 > 2017.12.31 16:25:03 LOG5[0]: s_connect: connected 173.194.67.108:465 > 2017.12.31 16:25:03 LOG5[0]: Service [gmail-smtp] connected remote server from 192.168.1.103:1142 > 2017.12.31 16:25:03 LOG5[0]: OCSP: Connecting the AIA responder "http://ocsp.pki.goog/gsr2" > 2017.12.31 16:25:04 LOG5[0]: s_connect: connected 172.217.14.174:80 > 2017.12.31 16:25:04 LOG5[0]: OCSP: Certificate accepted > 2017.12.31 16:25:04 LOG5[0]: OCSP: Connecting the AIA responder "http://ocsp.pki.goog/GTSGIAG3" > 2017.12.31 16:25:04 LOG5[0]: s_connect: connected 172.217.14.174:80 > 2017.12.31 16:25:04 LOG5[0]: OCSP: Certificate accepted > 2017.12.31 16:25:04 LOG5[0]: Certificate accepted at depth=0: C=US, ST=California, L=Mountain View, O=Google Inc, CN=smtp.gmail.com > 2017.12.31 16:25:04 LOG5[0]: Connection closed: 55 byte(s) sent to TLS, 514 byte(s) sent to socket > > > and this is what shows up on my monitoring program log file... > > To = xxxxxx(a)gmail.com > Connecting to mail server... 127.0.0.1 > Sending message... > E-mail Failed. Error: 20162 / 530 5.5.1 https://support.google.com/mail/?p=WantAuthError 51sm9558225oti.10 - gsmtp > 12/31/2017 4:25:04 PM > _______________________________________________ > stunnel-users mailing list > stunnel-users(a)stunnel.org > https://www.stunnel.org/cgi-bin/mailman/listinfo/stunnel-users

1 0

outgoing mail solution?
by Parker Brooks 02 Jan '18

02 Jan '18

hello and happy new year! it seemed this question has been asked before in various ways, however with the frequent version updates i'm still unclear if stunnel is meant as a fix to my issue. i have a outdated home monitoring program running on a winxp box. when an error occurs it sends an email. it's setup screen (besides the destination email field) only have a) smtp server, b) undeliverable email address and c) port # fields to configure. my workaround for several years (when port 25 became blocked or needed authetication) has been to use LOApost but they've recently ceased. that program, also running in the background, and if i understand correctly acted as a local host. i configured the monitoring program's smtp for 127.0.0.1 and LOApost did the rest other searches suggest cloud solutions like sendgrid and authsmtp. but those sites need a login/ password with the outgoing message. perhaps this is where i'm getting confused between the differences of SSL and smtp authentication? i have gotten gsuite's restricted smtp server to work, but i hoped stunnel would provide a better solution (ability to send besides gmail/ gsuite addresses). below are my conf and log files of my attempt. again, i can't tell if i've missed a step, or if this is not intended to work. PS: i've also tried to understand if this feature is on the todo list? tia, parker stunnel.conf [gmail-pop3] client = yes accept = 127.0.0.1:110 connect = pop.gmail.com:995 verifyChain = yes CAfile = ca-certs.pem checkHost = pop.gmail.com OCSPaia = yes [gmail-imap] client = yes accept = 127.0.0.1:143 connect = imap.gmail.com:993 verifyChain = yes CAfile = ca-certs.pem checkHost = imap.gmail.com OCSPaia = yes [gmail-smtp] client = yes accept = 127.0.0.1:25 connect = smtp.gmail.com:465 verifyChain = yes CAfile = ca-certs.pem checkHost = smtp.gmail.com OCSPaia = yes log file 2017.12.31 16:22:45 LOG5[main]: stunnel 5.44 on x86-pc-msvc-1500 platform 2017.12.31 16:22:45 LOG5[main]: Compiled/running with OpenSSL 1.0.2m-fips 2 Nov 2017 2017.12.31 16:22:45 LOG5[main]: Threading:WIN32 Sockets:SELECT,IPv6 TLS:ENGINE,FIPS,OCSP,PSK,SNI 2017.12.31 16:22:45 LOG5[main]: Reading configuration from file stunnel.conf 2017.12.31 16:22:45 LOG5[main]: UTF-8 byte order mark not detected 2017.12.31 16:22:45 LOG5[main]: FIPS mode disabled 2017.12.31 16:22:45 LOG5[main]: Configuration successful 2017.12.31 16:25:03 LOG5[0]: Service [gmail-smtp] accepted connection from 127.0.0.1:1141 2017.12.31 16:25:03 LOG5[0]: s_connect: connected 173.194.67.108:465 2017.12.31 16:25:03 LOG5[0]: Service [gmail-smtp] connected remote server from 192.168.1.103:1142 2017.12.31 16:25:03 LOG5[0]: OCSP: Connecting the AIA responder " http://ocsp.pki.goog/gsr2" 2017.12.31 16:25:04 LOG5[0]: s_connect: connected 172.217.14.174:80 2017.12.31 16:25:04 LOG5[0]: OCSP: Certificate accepted 2017.12.31 16:25:04 LOG5[0]: OCSP: Connecting the AIA responder " http://ocsp.pki.goog/GTSGIAG3" 2017.12.31 16:25:04 LOG5[0]: s_connect: connected 172.217.14.174:80 2017.12.31 16:25:04 LOG5[0]: OCSP: Certificate accepted 2017.12.31 16:25:04 LOG5[0]: Certificate accepted at depth=0: C=US, ST=California, L=Mountain View, O=Google Inc, CN=smtp.gmail.com 2017.12.31 16:25:04 LOG5[0]: Connection closed: 55 byte(s) sent to TLS, 514 byte(s) sent to socket and this is what shows up on my monitoring program log file... To = xxxxxx(a)gmail.com Connecting to mail server... 127.0.0.1 Sending message... E-mail Failed. Error: 20162 / 530 5.5.1 https://support.google.com/ mail/?p=WantAuthError 51sm9558225oti.10 - gsmtp 12/31/2017 4:25:04 PM

3 2

Updating stunnel 5.43 to 5.44 fails on Win7prof(32)
by tagtraeumer 15 Dec '17

15 Dec '17

Hello, removing compression = zlib did fix the starting issue for v5.44, update worked, thanks. Is 'deflate' affected too? Is compression recommended at this stage of mail processing? Thanks once more, regards! > Hello, > The compression is not recommended with TLS because of the CRIME or > BREACH attacks. > From stunnel v 5.44 in Windows the zlib compression is not supported. > So you should remove the option: > compression = zlib > > If you want to use this compression, please place the file zlib1.dll > (from the previous version) in the stunnel / bin directory. > > Best regards, > Malgorzata

1 0

Updating stunnel 5.43 to 5.44 fails on Win7prof(32)
by tagtraeumer 14 Dec '17

14 Dec '17

Hello, I did try updating stunnel 5.43 to 5.44 on a Win7prof (32bit) system, but it fails... Log before update (v5.43): 2017.12.14 20:46:39 LOG5[main]: stunnel 5.43 on x86-pc-msvc-1500 platform 2017.12.14 20:46:39 LOG5[main]: Compiled/running with OpenSSL 1.0.2m-fips 2 Nov 2017 2017.12.14 20:46:39 LOG5[main]: Threading:WIN32 Sockets:SELECT,IPv6 TLS:ENGINE,FIPS,OCSP,PSK,SNI 2017.12.14 20:46:39 LOG5[main]: Reading configuration from file stunnel.conf 2017.12.14 20:46:39 LOG5[main]: UTF-8 byte order mark detected 2017.12.14 20:46:39 LOG6[main]: UI not supported by engine #1 (capi) 2017.12.14 20:46:39 LOG6[main]: Engine #1 (capi) initialized 2017.12.14 20:46:39 LOG5[main]: FIPS mode disabled 2017.12.14 20:46:39 LOG6[main]: Compression enabled: 2 method(s) 2017.12.14 20:46:39 LOG6[main]: Initializing service [HostEurope_SMTPs] 2017.12.14 20:46:39 LOG6[main]: Loading certificate from file: stunnel.pem 2017.12.14 20:46:39 LOG6[main]: Certificate loaded from file: stunnel.pem 2017.12.14 20:46:39 LOG6[main]: Loading private key from file: stunnel.pem 2017.12.14 20:46:39 LOG6[main]: Private key loaded from file: stunnel.pem 2017.12.14 20:46:39 LOG4[main]: Service [HostEurope_SMTPs] uses "verifyChain" without subject checks 2017.12.14 20:46:39 LOG4[main]: Use "checkHost" or "checkIP" to restrict trusted certificates 2017.12.14 20:46:39 LOG5[main]: Configuration successful Log after 5.44 update: Pls refer to screenshot at: https://imgur.com/hCOKES7 "Invalid configuration file" errno: (*_errno()) stunnel.conf:65: ["name"]: Failed to initialize TLS How to solve this issue? Thx + regards!

1 0

Exception 40000015 STATUS_FATAL_APP_EXIT
by alexwaskilledfriendlyfire＠gmail.com 11 Dec '17

11 Dec '17

Hello Stunnel-users, Well, I have a problem with latest stunnel-5.44-win32 build. Both GUI and console versions won't to start throwing unhandled exception: tstunnel.exe: Problem signature: Problem Event Name: APPCRASH Application Name: tstunnel.exe Application Version: 0.0.0.0 Application Timestamp: 5a18f4f5 Fault Module Name: tstunnel.exe Fault Module Version: 0.0.0.0 Fault Module Timestamp: 5a18f4f5 Exception Code: 40000015 Exception Offset: 0000e43d OS Version: 6.1.7601.2.1.0.256.4 Locale ID: 1049 Additional Information 1: 57b4 Additional Information 2: 57b4bc5bbe66a065d3bc087357c09646 Additional Information 3: f6bb Additional Information 4: f6bb7b9b6fcf2c5cdb238bc389e9ede9 And GUI version stunnel.exe: --------------------------- Microsoft Visual C++ Runtime Library --------------------------- Runtime Error! Program: C:\Progs\Stunnel\bin\stunnel.exe This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information. --------------------------- OK --------------------------- Problem signature: Problem Event Name: APPCRASH Application Name: stunnel.exe Application Version: 5.44.0.0 Application Timestamp: 5a18f4f3 Fault Module Name: stunnel.exe Fault Module Version: 5.44.0.0 Fault Module Timestamp: 5a18f4f3 Exception Code: 40000015 Exception Offset: 0000e43d OS Version: 6.1.7601.2.1.0.256.4 Locale ID: 1049 Additional Information 1: 2de9 Additional Information 2: 2de98e0ccee9cc0b77f8fe3f51196ec8 Additional Information 3: 42db Additional Information 4: 42db6f5f1de1e71ee661917dbb7cc0c8 Completely fresh isntallation (or deleting config folder) doesn't solve my problem. My OS is legal Windows 7 x64 with latest updates. There are screenshots just for case. https://i.imgur.com/N1kPI9J.jpg https://i.imgur.com/CqMXWtx.jpg Probably I missing some DLL or service running but it's not an option to turn everything ON (and copy every single DLL ever made to my PC) to determine which one is not running. Probably it's something else. I need help. Thanks. -- Best regards, Alexwaskilledfriendlyfire mailto:alexwaskilledfriendlyfire@gmail.com

2 3

[patch] Fix build with LibreSSL
by Piotr Kubaj 06 Dec '17

06 Dec '17

Building against LibreSSL currently fails. Could you commit this patch? It builds fine on FreeBSD 11.1-RELEASE against LibreSSL 2.6.3. Since it doesn't change anything for OpenSSL, it should build against it too. -- ______________________________________ / A nickel ain't worth a dime anymore. \ | | \ -- Yogi Berra / -------------------------------------- \ ^__^ \ (oo)\_______ (__)\ )\/\ ||----w | || ||

1 0

possibility to delay connect to host
by Paavo Sild 04 Dec '17

04 Dec '17

Hello! I am a bit new to stunnel and while I tried to search the message archive I did not find any possible reference to the issue I try to solve. I am running a bit older version of stunnel as a simple reverse proxy. It accepts incoming TLS connection and opens clear TCP connection to a host specified by "connect" parameter in config. I wander - it is possible to delay the opening of TCP connection to a host until the actual data is received from client? I haven't found any reference to such a possibility in stunnel but thought that it worth to ask from community. The issue is that due to the nature of traffic there are instances then client negotiates the SSL connection but there is no actual data to be sent by the client. By opening the connection to host only when actual data is present would filter out unnecessary connections towards the application. Anybody had such an issue to solve? -- rgds, Paavo Sild --------------------- paavo.sild(a)estcard.ee ---------------------

1 0

routines:ssl3_read_bytes:sslv3 alert certificate unknown
by Ziad Badawi 04 Dec '17

04 Dec '17

Greetings, I am trying to capture clear text pcaps from client (browser) - server (java appserver) traffic. The java appserver is jboss using https. I'm running jboss and stunnel on the same machine. # stunnel.conf debug = 3 foreground = yes [jboss] client = yes cert= stunnel.pem # generated using makecert.sh accept = 1234 connect = 127.0.0.1:443 Version: stunnel 5.44 on x86_64-pc-linux-gnu platform Compiled/running with OpenSSL 1.0.2k-fips 26 Jan 2017 Threading:PTHREAD Sockets:POLL,IPv6 TLS:ENGINE,FIPS,OCSP,PSK,SNI Global options: RNDbytes = 64 RNDfile = /dev/urandom RNDoverwrite = yes Service-level options: ciphers = FIPS (with "fips = yes") ciphers = HIGH:!DH:!aNULL:!SSLv2 (with "fips = no") curve = prime256v1 debug = daemon.notice logId = sequential options = NO_SSLv2 options = NO_SSLv3 sessionCacheSize = 1000 sessionCacheTimeout = 300 seconds stack = 65536 bytes TIMEOUTbusy = 300 seconds TIMEOUTclose = 60 seconds TIMEOUTconnect = 10 seconds TIMEOUTidle = 43200 seconds verify = none When I try to test it usng firefox by browsing to https://localhost:1234, FF returns "Secure Connection Failed" and stunnel spits 2017.12.01 20:35:10 LOG3[0]: SSL_connect: 14094416: error:14094416:SSL routines:ssl3_read_bytes:sslv3 alert certificate unknown What am I missing / doing wrong? Regards Z

2 2

Stunnel 5.43 problem
by Carter Browne 15 Nov '17

15 Nov '17

When running stunnel 5.43 on Linux 4.14 (arm), accept = port only listens on 127.0.0.1:port. I don't know if other operating systems are affected. Carter

1 0

Re: [stunnel-users] TLS "translation" & 2-way auth
by Igor Gatis 15 Nov '17

15 Nov '17

It would be nice to know whether it is actually possible to achieve this with stunnel. If not, is there any other tool I could use or combine? On Nov 13, 2017 08:58, "Igor Gatis" <igorgatis(a)gmail.com> wrote: Yep, that's exactly what I'm seeking for help here. If we can abstract the 2-way bit for a second, I'd call this a "certificate transcription" TLS tunnel. On Thu, Nov 9, 2017 at 5:19 PM, Vincent Deschenes <vdeschenes(a)stelvio.com> wrote: > Ho, > > But that does not account for the A ->[TLS] ->B part. > > I believe that my sample will listen for unencrypted connection only. > > > > > > *From:* stunnel-users [mailto:stunnel-users-bounces@stunnel.org] *On > Behalf Of *Vincent Deschenes > *Sent:* Thursday, 9 November 2017 3:16 PM > *To:* Igor Gatis <igorgatis(a)gmail.com>; stunnel-users(a)stunnel.org > *Subject:* Re: [stunnel-users] TLS "translation" & 2-way auth > > > > You need to have a section in your config file which listen for requests > but also have the “client = yes” option with a cert and key like this: > > > > [http_a_to_c] > > client = yes > > accept = port_number_to_listen_on_server_b > > connect = server_c_address:443 > > cert = certificate.crt > > key = private.key > > > > > > cert and key are the certificate and private key server B uses to identify > itself on server C. > > You could also add more options to specify a trustore to specify which > cert coming from server C server B will trust, otherwise server B will > simply allow the connection. > > > > Good Luck > > > > > > *From:* stunnel-users [mailto:stunnel-users-bounces@stunnel.org > <stunnel-users-bounces(a)stunnel.org>] *On Behalf Of *Igor Gatis > *Sent:* Thursday, 9 November 2017 1:14 PM > *To:* stunnel-users(a)stunnel.org > *Subject:* [stunnel-users] TLS "translation" & 2-way auth > > > > Consider scenario below: > > > > Server A ==TLS==> Server B ==TLS+2WayAuth==> Server C > > > > Server A needs to connect to Server C through Server B which runs Stunnel. > Server C requires 2-way authentication. I have full control over Server A > and Server B and Server C belongs to a third-party. > > > > What does Stunnel config should look like? > > >

3 2

Re: [stunnel-users] Possible stunnel 5.43 bug ?
by Bezspam 15 Nov '17

15 Nov '17

Hello, The patch fixed the problem. Thank you very much! Best regards, Bez Spam Sent with [ProtonMail](https://protonmail.com) Secure Email. > -------- Original Message -------- > Subject: Re: [stunnel-users] Possible stunnel 5.43 bug ? > Local Time: November 15, 2017 7:15 AM > UTC Time: November 15, 2017 7:15 AM > From: Michal.Trojnara(a)stunnel.org > To: stunnel-users(a)stunnel.org <stunnel-users(a)stunnel.org> > > On 11/14/2017 01:33 PM, Bezspam wrote: > >> Is this a bug or expected behaviour? >> Best regards, >> Bez Spam >> >> Dear Bez, >> >> I confirm this is a bug. A patch that fixes it is: >> ftp://ftp.stunnel.org/stunnel/accept.patch >> >> There will be a bugfix release of stunnel later this week. >> >> Best regards, >> Mike > > --------------------------------------------------------------- > > stunnel-users mailing list > stunnel-users(a)stunnel.org > https://www.stunnel.org/cgi-bin/mailman/listinfo/stunnel-users

1 0

Ubuntu 16.04 problem
by Carter Browne 14 Nov '17

14 Nov '17

I have a very confusing problem with stunnel on Ubuntu. I have a configuration file that I have used in both Windows and Arch Linux. In the Arch version, I use a single file with all the certificates for the partner systems, in the Windows version I use a capath directory. I have tried both here and I always get the response "tlsv1 alert unknown ca". The beginning of the stunnel.conf file is below. Any hints would be appreciated. Carter cert = /etc/stunnel/xxxxxxx.pem key = /etc/stunnel/xxxxxxx.pem fips = no # Enable client mode client = yes ; Some security enhancements for UNIX systems - comment them out on Win32 #chroot = /var/run/stunnel/ #setuid = stunnel #setgid = stunnel ; PID is created inside chroot jail pid = /etc/stunnel/stunnel.pid ; Some performance tunings # Debug and Foreground are for testing socket = l:TCP_NODELAY=1 socket = r:TCP_NODELAY=1 # Uncomment for debugging stunnel problems debug = 5 output = /etc/stunnel/stunnel.log ##foreground = yes ; NO SSLv2 options = NO_SSLv3 sslVersion = all Ciphers = ECDHE-RSA-AES256-SHA:ECDHE-ECDSA-AES256-SHA:ECDH-RSA-AES256-SHA:ECDH-ECDSA-AES256-SHA:ECDHE-RSA-AES128-SHA:ECDHE-ECDSA-AES128-SHA:ECDH-RSA-AES128-SHA:ECDH-ECDSA-AES128-SHA # # Uncomment and modify the CAFile to point to your ca.txt file #CAFile = /etc/stunnel/all.pem CAPath = /etc/stunnel/capath #CAPath = @sysconfdir/ssl/certs # Uncomment and modify the verify to enable CA validation # verifyChain = no VerifyPeer = yes

1 0

Possible stunnel 5.43 bug ?
by Bezspam 14 Nov '17

14 Nov '17

Hello, I use stunnel on FreeBSD 11.1. I install it from ports. Yesterday I updated to stunnel 5.43. However it stopped working. I investigated the problem and found the issue- instead on all network interfaces, "accept" listens on 127.0.0.1 : sockstat -4 USER COMMAND PID FD PROTO LOCAL ADDRESS FOREIGN ADDRESS stunnel stunnel 43945 7 tcp4 127.0.0.1:2222 *:* stunnel stunnel 43945 8 tcp4 127.0.0.1:3128 *:* stunnel stunnel 43945 9 tcp4 127.0.0.1:8080 *:* stunnel stunnel 43945 10 tcp4 127.0.0.1:1616 *:* A workaround is to modify "accept = 0.0.0.0:3128" explicitly to listen on 0.0.0.0. Is this a bug or expected behaviour? Best regards, Bez Spam

2 1

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

stunnel-users