- stunnel-users - stunnel.org

High CPU system usage on Solaris
by armando.salvatori＠accenture.com 05 Mar '07

05 Mar '07

We use stunnel process to encrypt Oracle connections from a set of 4-6 application server. On the DB server, 20 cpus Solaris machine running Oracle, we have more then 400 LWPS for stunnel and the relative 400 oracle connections. We experience a huge amount of cpu usage (system component), about 40%, after stunnel setup (stunnel 4.15 version) Is it usual? Do you have experience on similar situation? Do you have any suggestions to improve cpu usage or to address performance issue/tuning? Thank you very much AS This message is for the designated recipient only and may contain privileged, proprietary, or otherwise private information. If you have received it in error, please notify the sender immediately and delete the original. Any other use of the email by you is prohibited.

1 0

make compile error options.c parse_ocsp_url
by William Biagi 03 Mar '07

03 Mar '07

./configure was successful but the make gave me the following: Making all in src make[1]: Entering directory `/home/bbiagi/stunnel-4.20/src' if gcc -DPACKAGE_NAME=\"stunnel\" -DPACKAGE_TARNAME=\"stunnel\" -DPACKAGE_VERSION=\"4.20\" -DPACKAGE_STRING=\"stunnel\ 4.20\" -DPACKAGE_BUGREPORT=\"\" -DPACKAGE=\"stunnel\" -DVERSION=\"4.20\" -DHOST=\"i686-pc-linux-gnu\" -DCPU_I686=1 -DVENDOR_PC=1 -DOS_LINUX_GNU=1 -DSTDC_HEADERS=1 -DHAVE_SYS_TYPES_H=1 -DHAVE_SYS_STAT_H=1 -DHAVE_STDLIB_H=1 -DHAVE_STRING_H=1 -DHAVE_MEMORY_H=1 -DHAVE_STRINGS_H=1 -DHAVE_INTTYPES_H=1 -DHAVE_STDINT_H=1 -DHAVE_UNISTD_H=1 -DHAVE_DLFCN_H=1 -DSIZEOF_UNSIGNED_CHAR=1 -DSIZEOF_UNSIGNED_SHORT=2 -DSIZEOF_UNSIGNED_INT=4 -DSIZEOF_UNSIGNED_LONG=4 -DHAVE_DEV_PTMX=1 -DRANDOM_FILE=\"/dev/urandom\" -DHAVE_UCONTEXT_H=1 -DHAVE_PTHREAD_H=1 -DHAVE_SYS_SELECT_H=1 -DHAVE_POLL_H=1 -DHAVE_SYS_POLL_H=1 -DHAVE_TCPD_H=1 -DHAVE_SYS_IOCTL_H=1 -DHAVE_STROPTS_H=1 -DHAVE_GRP_H=1 -DHAVE_UNISTD_H=1 -DHAVE_SYS_RESOURCE_H=1 -DHAVE_PTY_H=1 -DHAVE_LIBUTIL=1 -DHAVE_LIBDL=1 -DHAVE_LIBZ=1 -DHAVE_LIBPTHREAD=1 -DUSE_PTHREAD=1 -DHAVE_SNPRINTF=1 -DHAVE_VSNPRINTF=1 -DHAVE_OPENPTY=1 -DHAVE_DAEMON=1 -DHAVE_WAITPID=1 -DHAVE_WAIT4=1 -DHAVE_SETSID=1 -DHAVE_SETGROUPS=1 -DHAVE_CHROOT=1 -DHAVE_SYSCONF=1 -DHAVE_GETRLIMIT=1 -DHAVE_PTHREAD_SIGMASK=1 -DHAVE_LOCALTIME_R=1 -DHAVE_GETCONTEXT=1 -DHAVE_POLL=1 -DHAVE_ENDHOSTENT=1 -DHAVE_GETNAMEINFO=1 -DHAVE_GETADDRINFO=1 -DHAVE_OPENSSL=1 -Dssldir=\"/usr\" -DHAVE_OSSL_ENGINE_H=1 -I. -I. -I/usr/kerberos/include -DLIBDIR='"/usr/local/lib"' -DCONFDIR='"/usr/local/etc/stunnel"' -DPIDFILE='"/usr/local/var/run/stunnel/stunnel.pid"' -g -O2 -Wall -Wshadow -Wcast-align -Wpointer-arith -I/usr/include -MT options.o -MD -MP -MF ".deps/options.Tpo" -c -o options.o options.c; \ then mv -f ".deps/options.Tpo" ".deps/options.Po"; else rm -f ".deps/options.Tpo"; exit 1; fi options.c: In function `parse_ocsp_url': options.c:1762: warning: implicit declaration of function `OCSP_parse_url' options.c:1767: structure has no member named `ocsp_addr' options.c:1769: structure has no member named `ocsp_path' options.c: In function `parse_ocsp_flag': options.c:1786: `OCSP_NOCERTS' undeclared (first use in this function) options.c:1786: (Each undeclared identifier is reported only once options.c:1786: for each function it appears in.) options.c:1787: `OCSP_NOINTERN' undeclared (first use in this function) options.c:1788: `OCSP_NOSIGS' undeclared (first use in this function) options.c:1789: `OCSP_NOCHAIN' undeclared (first use in this function) options.c:1790: `OCSP_NOVERIFY' undeclared (first use in this function) options.c:1791: `OCSP_NOEXPLICIT' undeclared (first use in this function) options.c:1792: `OCSP_NOCASIGN' undeclared (first use in this function) options.c:1793: `OCSP_NODELEGATED' undeclared (first use in this function) options.c:1794: `OCSP_NOCHECKS' undeclared (first use in this function) options.c:1795: `OCSP_TRUSTOTHER' undeclared (first use in this function) options.c:1796: `OCSP_RESPID_KEY' undeclared (first use in this function) options.c:1797: `OCSP_NOTIME' undeclared (first use in this function) options.c: At top level: options.c:1758: warning: `parse_ocsp_url' defined but not used options.c:1781: warning: `parse_ocsp_flag' defined but not used make[1]: *** [options.o] Error 1 make[1]: Leaving directory `/home/bbiagi/stunnel-4.20/src' make: *** [all-recursive] Error 1 I am trying to install on a red hat 7.3 with openssl-0.9.6b-18

2 1

RE: Stunnel Forks Out of Control
by Duncan Brown 24 Feb '07

24 Feb '07

Hi all, Ok, I've gone through my last log file for when this issue happens. There's "1130 finished (2 left)" (where 1030 is the port we're listening on) at 10:22am, for example. Then, the next "1130 finished" doesn't happen for about another 20 minutes (which is weird), and then says, "1131 finished (124 left)". It's at this point that a lot of "connection reset by peer"'s happen and "0 bytes sent to SSL" also happen. It looks as though there are a bunch of connections being accepted but not handled. Can anyone assist? The debug level 7 log file is available if necessary. Thanks again, guys. Cheers, Duncan

2 1

RE: Stunnel Forks Out of Control
by Duncan Brown 23 Feb '07

23 Feb '07

Hi all, As an update, it looks like I'm using Stunnel 4.15. I thought it said 4.20, unless I managed to compile an older version. Regardless, I think the problem is happening in both versions. It looks as though it can't catch up on processing cached requests or something. Does that help at all? Cheers, Duncan

1 0

Stunnel Forks Out of Control
by Duncan Brown 21 Feb '07

21 Feb '07

Hi all, I apologize if this issue appears in the archives but I've scoured them and can't find any similar instances (if this problem does occur in the archives then if someone could point it out to me I'd be grateful as well). So, here goes: I'm running Stunnel on Debian Linux. I run it in daemon mode and have the .conf file set to listen on a specific port and redirect to another port. Listening on the connecting end of Stunnel is another daemon I've created that listens on the same port for TCP traffic as part of a client-server application I have (a new process is forked for each connection). On the client end is a .NET application that communicates with the server via TCP with Stunnel installed. In my code, I've made sure that whenever a client-server session is finished, the server closes the socket, whether or not there was an error. Now, everything seems to work flawlessly for a period of time (this period of time is variable, from anywhere between 1 day to 1 month). Then, all of a sudden, Stunnel "explodes"; for example, while there might be one or two forks of my own daemon, there might be 128 instances of Stunnel, all of which are in a sleeping state (that is, if I run ps -aux, their "STAT" attribute is "S"), whereas before it exploded it was pretty much a 1:1 relationship. When it "explodes", I have to kill Stunnel and restart it, which is a pain in the rear-end given what I'm trying to do. Why is it doing this? How do I stop it from happening? I apologize again if this has already been brought up somewhere else, but I've been unable to find any references to any similar problems. If it's a "newbie" mistake or something I'll gladly accept it and learn from it. Please find below the information requested for a problem for the mailing list: - I'm using v4.20 - No parameters are used from the command-line when starting Stunnel; it just uses the standard .conf file which I've edited. - Output of "uname -a" : "Linux buffed 2.4.27-2-386 #1 Wed Aug 17 09:33:35 UTC 2005 i686 GNU/Linux" - I'm using libc v6 - Output of "gcc -v" : Reading specs from /usr/lib/gcc-lib/i486-linux/3.3.5/specs Configured with: ../src/configure -v --enable-languages=c,c++,java,f77,pascal,objc,ada,treelang --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-gxx-include-dir=/usr/include/c++/3.3 --enable-shared --enable-__cxa_atexit --with-system-zlib --enable-nls --without-included-gettext --enable-clocale=gnu --enable-debug --enable-java-gc=boehm --enable-java-awt=xlib --enable-objc-gc i486-linux Thread model: posix gcc version 3.3.5 (Debian 1:3.3.5-13) - I'm using OpenSSL v0.9.7e (25 Oct 2004) Thanks in advance for any help. Cheers, Duncan

1 0

Stunnel proxy aware
by Kennet Svanberg 21 Feb '07

21 Feb '07

Hi, I'm quite happy with stunnel-4.05 but don't know how to add any of the proxy patches to make my own windows binaries. I also saw that others have wondered http://linode.mirt.net/pipermail/stunnel-users/2006-December/001404.html

1 0

Stunnel on AIX 5.3 and AIX 5.2 - Solution
by Trevor Maistry 21 Feb '07

21 Feb '07

Hi, I downloaded stunnel version 4.05, changed the configuration file to use "cc" instead of "gcc" and compiled on AIX 5300-04 and AIX 5200-03. Compilations worked fine. However, you must compile on the machine you want to use it on. Otherwise you will have problems. Copy the "stunnel", "stunnel.pem" and sample "stunnel.conf-sample.in" files into /usr/local/bin. Rename the "stunnel.conf-sample.in" to " "stunnel.conf". Change the "stunnel.conf" accordingly and start "stunnel" like so "./stunnel stunnel.conf". Hope this helps. Regards, Trevor Maistry Principal Systems Consultant ============================================= Information and Communication Technology Division University of KwaZulu-Natal King George v Avenue, Durban, 4041 Tel : +2731-2602475 Fax: +2731-2602125 Cell: +2782-787-5583 Email : maistryt(a)ukzn.ac.za =============================================: Please find our Email Disclaimer here: http://www.ukzn.ac.za/disclaimer/

1 0

Re: new Cygwin maintainer for stunnel
by Michal Trojnara 20 Feb '07

20 Feb '07

On Tuesday 20 February 2007 20:59, Andrew Schulman wrote: > Please add Cygwin to the list of "known good" OSes at > http://stunnel.mirt.net/, and add me to the list of maintainers in the > PORTS file. Done. Thank you. Mike

1 0

help
by RICH 16 Feb '07

16 Feb '07

Hi to all i need help i have install stunnel rpm in mandriva but can not find the config file

1 0

stunnel on Qnx 6.3
by Rajashekhar Durgadmath 15 Feb '07

15 Feb '07

Hello All, I am facing problems running stunnel for Qnx 6.3 cross compiled on solaris. I want to provide a https front end to thtttpd (www.acme.com) stunnel starts but fails. Logs are like this: 1980.01.01 15:01:46 LOG5[1069085:1]: stunnel 4.20 on sparc-sun-solaris2.9 with OpenSSL 0.9.8 05 Jul 2005 1980.01.01 15:01:46 LOG5[1069085:1]: Threading:PTHREAD SSL:ENGINE Sockets:POLL,IPv4 1980.01.01 15:01:46 LOG6[1069085:1]: file ulimit = 1000 (can be changed with 'ulimit -n') 1980.01.01 15:01:46 LOG6[1069085:1]: poll() used - no FD_SETSIZE limit for file descriptors 1980.01.01 15:01:46 LOG5[1069085:1]: 488 clients allowed 1980.01.01 15:01:46 LOG7[1069085:1]: FD 4 in non-blocking mode 1980.01.01 15:01:46 LOG7[1069085:1]: FD 5 in non-blocking mode 1980.01.01 15:01:46 LOG7[1069085:1]: FD 6 in non-blocking mode 1980.01.01 15:01:46 LOG7[1069085:1]: SO_REUSEADDR option set on accept socket 1980.01.01 15:01:46 LOG7[1069085:1]: stunnel bound to xxx.xx.xxx.xxx:443 1980.01.01 15:01:46 LOG7[1073200:1]: Created pid file /stunnel.pid <Browser connects> 1980.01.01 15:05:12 LOG7[1073200:1]: stunnel accepted FD=7 from yyy.yy.yyy.yyy:64822 1980.01.01 15:05:12 LOG7[1073200:2]: stunnel started 1980.01.01 15:05:12 LOG7[1073200:2]: FD 7 in non-blocking mode 1980.01.01 15:05:12 LOG7[1073200:2]: TCP_NODELAY option set on local socket 1980.01.01 15:05:12 LOG5[1073200:2]: stunnel accepted connection from qqq.qqq.qqq.qqq:64822 1980.01.01 15:05:12 LOG5[1073200:2]: Server mode 1980.01.01 15:05:12 LOG7[1073200:2]: SSL state (accept): before/accept initialization 1980.01.01 15:05:12 LOG7[1073200:2]: SSL state (accept): SSLv3 read client hello A 1980.01.01 15:05:12 LOG7[1073200:2]: SSL state (accept): SSLv3 write server hello A 1980.01.01 15:05:12 LOG7[1073200:2]: SSL state (accept): SSLv3 write certificate A 1980.01.01 15:05:12 LOG7[1073200:2]: SSL state (accept): SSLv3 write server done A 1980.01.01 15:05:12 LOG7[1073200:2]: SSL state (accept): SSLv3 flush data 1980.01.01 15:05:12 LOG7[1073200:2]: SSL state (accept): SSLv3 read client key exchange A 1980.01.01 15:05:13 LOG7[1073200:2]: SSL state (accept): SSLv3 read finished A 1980.01.01 15:05:13 LOG7[1073200:2]: SSL state (accept): SSLv3 write change cipher spec A 1980.01.01 15:05:13 LOG7[1073200:2]: SSL state (accept): SSLv3 write finished A 1980.01.01 15:05:13 LOG7[1073200:2]: SSL state (accept): SSLv3 flush data 1980.01.01 15:05:13 LOG7[1073200:2]: 1 items in the session cache 1980.01.01 15:05:13 LOG7[1073200:2]: 0 client connects (SSL_connect()) 1980.01.01 15:05:13 LOG7[1073200:2]: 0 client connects that finished 1980.01.01 15:05:13 LOG7[1073200:2]: 0 client renegotiations requested 1980.01.01 15:05:13 LOG7[1073200:2]: 1 server connects (SSL_accept()) 1980.01.01 15:05:13 LOG7[1073200:2]: 1 server connects that finished 1980.01.01 15:05:13 LOG7[1073200:2]: 0 server renegotiations requested 1980.01.01 15:05:13 LOG7[1073200:2]: 0 session cache hits 1980.01.01 15:05:13 LOG7[1073200:2]: 1 session cache misses 1980.01.01 15:05:13 LOG7[1073200:2]: 0 session cache timeouts 1980.01.01 15:05:13 LOG6[1073200:2]: SSL accepted: new session negotiated 1980.01.01 15:05:13 LOG6[1073200:2]: Negotiated ciphers: RC4-MD5 SSLv3 Kx=RSA Au=RSA Enc=RC4(128) Mac=MD5 < SSL Negotiation done > 1980.01.01 15:08:49 LOG3[1073200:2]: remote socket: Address family not supported by protocol family (247) <<<<<< 1980.01.01 15:08:49 LOG5[1073200:2]: Connection reset: 0 bytes sent to SSL, 0 bytes sent to socket 1980.01.01 15:08:49 LOG7[1073200:2]: stunnel finished (0 left) I checked the address family; it is AF_INET :) socket syscall in client.c:989 is failing. I know the error "Address family not supported by protocol family (247)" says it all, but there are other applications running on the box which do the same thing and it works for them. Versions: OpenSSL 0.9.8 and stunnel 4.20. stunnel.conf ## stunnel.conf starts cert = /etc/localhost_cert.pem key = /etc/localhost_key.pem ; Protocol version (all, SSLv2, SSLv3, TLSv1) sslVersion = all ; Some security enhancements for UNIX systems - comment them out on Win32 chroot = /some_dir setuid = root setgid = root ; PID is created inside chroot jail pid = /stunnel.pid ; Some performance tunings socket = l:TCP_NODELAY=1 socket = r:TCP_NODELAY=1 debug = 7 output = /some_dir/stunnel.log [https] accept = <Qnx box IP>:443 connect = <Qnx box IP>:80 ## stunnel.conf end # uname -a QNX localhost 6.3.0 2006/04/27-13:08:16EST armbe It was configured thus: ./configure --prefix=/vob/nmi/3rd_party/stunnel --enable-dependency-tracking --with-ssl=/vob/nmi/3rd_party/openssl/ --build=armbe-qnx --host=sparc-sun-solaris2.9 --with-threads=pthread --disable-libwrap Removing the --disable-libwrap does not help. stunnel is started like this: stunnel /etc/stunnel.conf -d http -r localhost:http -p /etc This also did not help: stunnel /etc/stunnel.conf -d https -r localhost:http -p /etc stunnel /etc/stunnel.conf # stunnel -version stunnel 4.20 on sparc-sun-solaris2.9 with OpenSSL 0.9.8 05 Jul 2005 Threading:PTHREAD SSL:ENGINE Sockets:POLL,IPv4 Global options debug = 5 pid = /some_path/stunnel.pid RNDbytes = 64 RNDfile = /dev/urandom RNDoverwrite = yes Service-level options cert = /some_path/stunnel.pem ciphers = ALL:!ADH:+RC4:@STRENGTH key = /some_path/stunnel.pem session = 300 seconds sslVersion = SSLv3 for client, all for server TIMEOUTbusy = 300 seconds TIMEOUTclose = 60 seconds TIMEOUTconnect = 10 seconds TIMEOUTidle = 43200 seconds verify = none Please reply, All comments appreciated. Cheers, Raj --------------------------------- New Yahoo! Mail is the ultimate force in competitive emailing. Find out more at the Yahoo! Mail Championships. Plus: play games and win prizes.

1 0