- stunnel-users - stunnel.org

Odd complication between VM clients, same host
by afpteam＠sbcglobal.net 24 Aug '14

24 Aug '14

Hi, Much appreciation to Stunnel and OpenSSL for a long, successful provision. I can't be sure yet if the concern below is Stunnel or OpenSSL related. I've surfed Google and found nothing similar. Stunnel package with integral OpenSSL is current release. I recently had to secure and connect two public facing client machines on the same Citrix Xen 6.2 release, running Server 2003 x 64 bit. I typically use Stunnel with self generated Pem, for securing short term TightVNC sessions. The config is always fairly generic, loopback allowed, securing across 127.0.0.1 Typically I don't make name references in the .conf profile, just accept / connect on related ports. This is the first time I recall ever pointing two VM's at each other. I have tried both service and gui on both Stunnel and VNC. Tried both application listing and discrete port listing in the server's windows firewall. I have tried updating security to the Admin account as owner or Administrators, in general. I have tried launching either or both as Administrator. When VNC client viewer calls to Server, Stunnel sees the connect, succeeds the Pem and stops. If I restart the stunnel service at the server, the viewer auth dialog suddenly succeeds. I question if this remains secured but clearly isn't practical for application. Both VM's are connected to me via RDP at the time, if this might relate to an Auth violation. I tried multiple VMs to eliminate anything related to a bungled VM config. I'm guessing there is some kind of circular routing restriction possibly. Citrix does use an X-window console to each VM on the control path, but in other instances there's never been any conflict. Not sure how to determine if the issue is config, Stunnel or OpenSSL or how to resolve it. Any suggestions would be appreciated. Regards and thank you again, Mike

1 0

Server side SNI and Windows problem
by Slava 17 Aug '14

17 Aug '14

I'm trying use SNI on Windows (server 2003 R2 or XP) stunnel 5.01 on x86-pc-msvc-1500 platform Compiled/running with OpenSSL 1.0.1g-fips 7 Apr 2014 stunnel.config ... [main] accept = 60000 connect = 10.0.21.1:80 [ssh] connect = 10.0.21.1:22 sni=main:ssh Client has line sni=ssh, but server not see this and attempt connect to :80. I'm try connect from linux and windows, same result For server on linux all work fine. ?

3 4

verify=4
by oweijr oweirj 13 Aug '14

13 Aug '14

Hello, In the stunnel documentation, I see the following: level 4 Ignore CA chain and only verify peer certificate. My interpretation of level 4 was that only the server certificate had to be installed on the client in order for the cert verification to pass. No issuer/CA certificates were needed. However, when I do this, the connection fails. I see this following in the logs with verify=4: 2014.08.12 21:17:57 LOG7[26768]: Starting certificate verification: depth=0, subject=/C=US/CN=a.b.com 2014.08.12 21:17:57 LOG4[26768]: CERT: Verification error: unable to get local issuer certificate 2014.08.12 21:17:57 LOG4[26768]: Certificate check failed: depth=0, subject=/C=US/CN=a.b.com 2014.08.12 21:17:57 LOG7[26768]: SSL alert (write): fatal: unknown CA 2014.08.12 21:17:57 LOG3[26768]: SSL_connect: 14090086: error:14090086:SSL routines:SSL3_GET_SERVER_CERTIFICATE:certificate verify failed My config file contains the following: verify = 4 CAfile = /opt/stunnel/certs/clients.pem chroot = /opt/stunnel setuid = stunnel setgid = stunnel pid = /run/stunnel.pid debug = 7 output = stunnel.log options = NO_SSLv2 cert = /opt/stunnel/certs/server.pem key = /opt/stunnel/certs/server.pem If I append the issuer certificate to the CAfile, after the peer certificate, then the connection is successful with verify=4. Is this expected behavior? Is there a verify level that works as I described above: only the peer certificate needs to be present, no CA/issuer certificates at all? Thank you for your help.

1 0

unable to send SMTP over SSL
by Paul Hargrove 09 Aug '14

09 Aug '14

Hi, I'm a new user using stunnel-5.03 on Windows SBS 2003. I can't send email from Exchange but can from Outlook with same credentials on same network. Email gets stuck in the exchange smtp queue. Simple config: ; * Global options debug = 7 output = stunnel.log engine = capi options = NO_SSLv2 ; * Service definition [SMTP] client = yes accept = 8465 connect = smtp.virginmedia.com:465 protocol = smtp Logs look good but nothing seems to go: 2014.08.08 22:15:07 LOG7[7284]: Dispatching signals from the signal pipe 2014.08.08 22:15:07 LOG7[7284]: Processing SIGNAL_RELOAD_CONFIG 2014.08.08 22:15:07 LOG5[7284]: Reading configuration from file stunnel.conf 2014.08.08 22:15:07 LOG7[7284]: Enabling support for engine 'capi' 2014.08.08 22:15:07 LOG7[7284]: Initializing engine #1 (capi) 2014.08.08 22:15:07 LOG6[7284]: Engine #1 (capi) initialized 2014.08.08 22:15:07 LOG5[7284]: FIPS mode disabled 2014.08.08 22:15:07 LOG7[7284]: Compression disabled 2014.08.08 22:15:07 LOG7[7284]: Snagged 64 random bytes from C:/.rnd 2014.08.08 22:15:07 LOG7[7284]: Wrote 1024 new random bytes to C:/.rnd 2014.08.08 22:15:07 LOG7[7284]: PRNG seeded successfully 2014.08.08 22:15:07 LOG6[7284]: Initializing service [SMTP] 2014.08.08 22:15:08 LOG7[7284]: No private key specified 2014.08.08 22:15:08 LOG7[7284]: SSL options set: 0x01000004 2014.08.08 22:15:08 LOG5[7284]: Configuration successful 2014.08.08 22:15:08 LOG7[7284]: Closing service [SMTP] 2014.08.08 22:15:08 LOG7[7284]: Service [SMTP] closed (FD=292) 2014.08.08 22:15:08 LOG7[7284]: Sessions cached before flush: 7 2014.08.08 22:15:08 LOG7[7284]: Sessions cached after flush: 0 2014.08.08 22:15:08 LOG7[7284]: Service [SMTP] closed 2014.08.08 22:15:08 LOG7[7284]: Service [SMTP] (FD=292) bound to 0.0.0.0:8465 2014.08.08 22:15:08 LOG7[7284]: Signal pipe is empty 2014.08.08 22:16:03 LOG7[7284]: Service [SMTP] accepted (FD=392) from 127.0.0.1:26642 2014.08.08 22:16:03 LOG7[7284]: Creating a new thread 2014.08.08 22:16:03 LOG7[7284]: New thread created 2014.08.08 22:16:03 LOG7[7184]: Service [SMTP] started 2014.08.08 22:16:03 LOG5[7184]: Service [SMTP] accepted connection from 127.0.0.1:26642 2014.08.08 22:16:03 LOG6[7184]: s_connect: connecting 62.254.26.219:465 2014.08.08 22:16:03 LOG7[7184]: s_connect: s_poll_wait 62.254.26.219:465: waiting 10 seconds 2014.08.08 22:16:03 LOG5[7184]: s_connect: connected 62.254.26.219:465 2014.08.08 22:16:03 LOG5[7184]: Service [SMTP] connected remote server from 10.0.0.2:26643 2014.08.08 22:16:03 LOG7[7184]: Remote socket (FD=688) initialized 2014.08.08 22:16:14 LOG3[7184]: fd_getline: Unexpected socket close 2014.08.08 22:16:14 LOG5[7184]: Connection reset: 0 byte(s) sent to SSL, 0 byte(s) sent to socket 2014.08.08 22:16:14 LOG7[7184]: Remote socket (FD=688) closed 2014.08.08 22:16:14 LOG7[7184]: Local socket (FD=392) closed 2014.08.08 22:16:14 LOG7[7184]: Service [SMTP] finished (0 left) 2014.08.08 22:16:14 LOG7[7184]: str_stats: 3 block(s), 4294966138 data byte(s), 150 control byte(s) 2014.08.08 22:16:14 LOG7[7184]: str_stats: 20 byte(s) at ..\src\network.c:413 2014.08.08 22:16:14 LOG7[7184]: str_stats: 20 byte(s) at ..\src\network.c:412 2014.08.08 22:16:14 LOG7[7184]: str_stats: 20 byte(s) at ..\src\network.c:411 If I rem out the protocol = smtp then the mails go from the queue and the logs seem to indicate a success: 2014.08.08 22:02:14 LOG7[7284]: Service [SMTP] accepted (FD=688) from 127.0.0.1:54520 2014.08.08 22:02:14 LOG7[7284]: Creating a new thread 2014.08.08 22:02:14 LOG7[7284]: New thread created 2014.08.08 22:02:14 LOG7[10072]: Service [SMTP] started 2014.08.08 22:02:14 LOG5[10072]: Service [SMTP] accepted connection from 127.0.0.1:54520 2014.08.08 22:02:14 LOG6[10072]: s_connect: connecting 62.254.26.219:465 2014.08.08 22:02:14 LOG7[10072]: s_connect: s_poll_wait 62.254.26.219:465: waiting 10 seconds 2014.08.08 22:02:14 LOG5[10072]: s_connect: connected 62.254.26.219:465 2014.08.08 22:02:14 LOG5[10072]: Service [SMTP] connected remote server from 10.0.0.2:54521 2014.08.08 22:02:14 LOG7[10072]: Remote socket (FD=456) initialized 2014.08.08 22:02:14 LOG6[10072]: SNI: sending servername: smtp.virginmedia.com 2014.08.08 22:02:14 LOG7[10072]: SSL state (connect): before/connect initialization 2014.08.08 22:02:14 LOG7[10072]: SSL state (connect): SSLv3 write client hello A 2014.08.08 22:02:14 LOG7[10072]: SSL state (connect): SSLv3 read server hello A 2014.08.08 22:02:14 LOG7[10072]: SSL state (connect): SSLv3 read finished A 2014.08.08 22:02:14 LOG7[10072]: SSL state (connect): SSLv3 write change cipher spec A 2014.08.08 22:02:14 LOG7[10072]: SSL state (connect): SSLv3 write finished A 2014.08.08 22:02:14 LOG7[10072]: SSL state (connect): SSLv3 flush data 2014.08.08 22:02:14 LOG7[10072]: 6 items in the session cache 2014.08.08 22:02:14 LOG7[10072]: 13 client connects (SSL_connect()) 2014.08.08 22:02:14 LOG7[10072]: 13 client connects that finished 2014.08.08 22:02:14 LOG7[10072]: 0 client renegotiations requested 2014.08.08 22:02:14 LOG7[10072]: 0 server connects (SSL_accept()) 2014.08.08 22:02:14 LOG7[10072]: 0 server connects that finished 2014.08.08 22:02:14 LOG7[10072]: 0 server renegotiations requested 2014.08.08 22:02:14 LOG7[10072]: 7 session cache hits 2014.08.08 22:02:14 LOG7[10072]: 0 external session cache hits 2014.08.08 22:02:14 LOG7[10072]: 0 session cache misses 2014.08.08 22:02:14 LOG7[10072]: 0 session cache timeouts 2014.08.08 22:02:14 LOG6[10072]: SSL connected: previous session reused 2014.08.08 22:02:16 LOG6[10072]: Read socket closed (readsocket) 2014.08.08 22:02:16 LOG7[10072]: Sending close_notify alert 2014.08.08 22:02:16 LOG7[10072]: SSL alert (write): warning: close notify 2014.08.08 22:02:16 LOG6[10072]: SSL_shutdown successfully sent close_notify alert 2014.08.08 22:02:16 LOG6[10072]: SSL socket closed (SSL_read) 2014.08.08 22:02:16 LOG7[10072]: Sent socket write shutdown 2014.08.08 22:02:16 LOG5[10072]: Connection closed: 80 byte(s) sent to SSL, 298 byte(s) sent to socket 2014.08.08 22:02:16 LOG7[10072]: Remote socket (FD=456) closed 2014.08.08 22:02:16 LOG7[10072]: Local socket (FD=688) closed 2014.08.08 22:02:16 LOG7[10072]: Service [SMTP] finished (0 left) 2014.08.08 22:02:16 LOG7[10072]: str_stats: 3 block(s), 4294963672 data byte(s), 150 control byte(s) 2014.08.08 22:02:16 LOG7[10072]: str_stats: 20 byte(s) at ..\src\network.c:413 2014.08.08 22:02:16 LOG7[10072]: str_stats: 20 byte(s) at ..\src\network.c:412 2014.08.08 22:02:16 LOG7[10072]: str_stats: 20 byte(s) at ..\src\network.c:411 but all Mails bounce with: There was a SMTP communication problem with the recipient's email server. Please contact your system administrator. <server.domain.local #5.5.0 smtp;530 Authentication Required (VM401)> Any ideas?

1 0

Re: [stunnel-users] stunnel 5.03 released
by Michal Trojnara 09 Aug '14

09 Aug '14

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Peter Pentchev wrote: > Um, there's just one little problem with this; what do you think > about the attached public-domain patch that removes what looks very > much like a stray quote character? Argh! Thank you. Mike -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iEYEARECAAYFAlPlP0MACgkQ/NU+nXTHMtG61QCgrh8w889DGai/Xbc7y42vTkQ/ /BQAnRsl60djjKX2O3PJWzRrVkLXDnEt =yKfI -----END PGP SIGNATURE-----

1 0

Re: [stunnel-users] stunnel 5.03 released
by Peter Pentchev 08 Aug '14

08 Aug '14

On Thu, Aug 07, 2014 at 10:15:47PM +0200, Michal Trojnara wrote: > Dear Users, > > I have released version 5.03 of stunnel. Hi, (-announce and openssl-users dropped from CC) Thanks a lot - and thanks for saving downstream distro maintainers the trouble of releasing an intermediate package update for the init script fix ;) > > The ChangeLog entry: [snip] > * Bugfixes [snip] > - Fixed whitespace handling in the stunnel.init script. Um, there's just one little problem with this; what do you think about the attached public-domain patch that removes what looks very much like a stray quote character? Once again, thanks for writing and maintaining stunnel! G'luck, Peter PS. I tried to send this from my <roam(a)ringlet.net> address, but mx1.mirt.net bounced my message due to a junkemailfilter.com blacklist problem that I've taken up with the folks over there (at junkemailfilter.com, not mirt.net, obviously ;) -- Peter Pentchev roam(a)ringlet.net roam(a)FreeBSD.org p.penchev(a)storpool.com PGP key: http://people.FreeBSD.org/~roam/roam.key.asc Key fingerprint 2EE7 A7A5 17FC 124C F115 C354 651E EFB0 2527 DF13

1 0

Windows Certificate Store
by j nasty 08 Aug '14

08 Aug '14

I haven't been able to get a definitive answer and can't find the information I need through traditional searching around. Anyways, is it possible to have sTunnel access the Windows certificate store to get the certificate that will be used for client authentication? If so, how can this be configured in sTunnel's configuration file? Or can you at least point me in the direction I can find the information I'm looking for? Many thanks in advance everyone!

2 1

stunnel 5.03 released
by Michal Trojnara 08 Aug '14

08 Aug '14

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Dear Users, I have released version 5.03 of stunnel. The ChangeLog entry: Version 5.03, 2014.08.07, urgency: HIGH: * Security bugfixes - OpenSSL DLLs updated to version 1.0.1i. See https://www.openssl.org/news/secadv_20140806.txt * New features - FIPS autoconfiguration cleanup. - FIPS canister updated to version 2.0.6. - Improved SNI diagnostic logging. * Bugfixes - Compilation fixes for old versions of OpenSSL. - Fixed whitespace handling in the stunnel.init script. Home page: https://www.stunnel.org/ Download: https://www.stunnel.org/downloads.html SHA-256 hashes: 5e8196f913bf7460082c1c7e5d436fbfba7f65d56c60065e6ccf0df9057294ef stunnel-5.03-android.zip 803bc85fbc80f22d71b13d4180e2d51c02586b2cc611169961a2745e47e31c4f stunnel-5.03-installer.exe 9a1e369466fa756e6f48b11480a3338c1fa4717e6472871bf4a3a96c483edd03 stunnel-5.03.tar.gz Best regards, Mike -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iEYEARECAAYFAlPj3nMACgkQ/NU+nXTHMtGM3ACfdFm7k754Pzo1rhDkVb6rT5nO EEUAoJoSXbRPHu3jphxXYVA5SeJ05BXj =tFws -----END PGP SIGNATURE-----

1 0

Re: [stunnel-users] cert verification problem
by n n 07 Aug '14

07 Aug '14

Hi! I want to authenticate server, using valid cert from StartSSL, and I receive error: ## sie 07 16:24:17 yabba stunnel[4247]: LOG4[5495]: CERT: Verification error: unable to get local issuer certificate sie 07 16:24:17 yabba stunnel[4247]: LOG4[5495]: Certificate check failed: depth=1, subject=/C=IL/O=StartCom Ltd./OU=Secure Digital Certificate Signing/CN=StartCom Class 1 Primary Intermediate Server CA sie 07 16:24:17 yabba stunnel[4247]: LOG3[5495]: SSL_connect: 14090086: error:14090086:SSL routines:SSL3_GET_SERVER_CERTIFICATE:certificate verify failed sie 07 16:24:17 yabba stunnel[4247]: LOG5[5495]: Connection reset: 0 byte(s) sent to SSL, 0 byte(s) sent to socket ## I don't know what's wrong. There is server's config: ## chroot = /var/run/stunnel/ pid=/stunnel.pid setuid = nobody setgid = nobody [squid] accept = 80 connect = 127.0.0.1:3128 verify=0 ## Client's config: ## client = yes [proxy] accept = 127.0.0.1:8005 connect = server:80 ##

1 0

Possible false positive trojan detection on stunnel-5.02-installer.exe
by Drew Mayo 07 Aug '14

07 Aug '14

Hi all, This is just an FYI. I'm a recent user of stunnel, and one of my team has just alerted me that f-prot (and unfortunately hence google drive) detects the current windows installer for stunnel 5.02 as a trojan. virustotal (which confirms it matches the sha256) tells me that the file is detected as: W32/Trojan.TDWH-2255 W32/Trojan3.JUA I've raised a misclassification request with f-prot, but of course that will probably take some time to get resolved. Of course, it is remotely possible it's not false, but i'm betting the other way. -d

2 2