- stunnel-users - stunnel.org

too many file descriptors
by Sligar, Benjamin 22 Apr '06

22 Apr '06

I'm working on upgrading from stunnel 4.04 to the 4.15 to utilize the per port specification of a certificate. On my initial testing, I ran with only a few ports, specified the different certs, and, everything seems to work fine. I'm ran into a problem, though, when trying to use the same config file in regards to the number of configured ports. I have 104 inbound ports setup, and, when it hits the 69th configured port, I get the following message for this, and every port that follows: s_poll_add failed for FD=69: too many file descriptors I checked the code, and, this is what I found: #define MAX_FD 64 I didn't find a relation to any configurable options, so I made a change to this parameter in the code, recompiled, and, I was able to start up all of my listens without error. I just wanted to pass this along, and, if it does tie into a configurable parameter, my apologies for not finding it....if someone knows it, please pass that on. Thanks! **************************************** Benjamin Sligar Senior Development Support Specialist Transaction Network Services, Inc P: 703.453.8324 F: 703.453.8391 AIM: dagdromer88 **************************************** This e-mail message is for the sole use of the intended recipient(s) and may contain confidential and privileged information of Transaction Network Services. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message.

2 1

Stunnel with TPM engine !!!
by dinesh.kallath＠bt.com 19 Apr '06

19 Apr '06

Hello I am new to Stunnel and would like to know how to integrate the engine with Stunnel 4.15. I am trying to integrate a Trusted Platform Module (TPM) engine which is compatible with OpenSSL to use with Stunnel so that the private key for SSL connection can be retrieved and stored in the hardware. I was able to configure the Stunnel config file to use the engine and it is loading the engine fine. The problem I am facing now is the key which stunnel tries to load can be loaded only by the engine, I mean it has to be loaded into the TPM to use it and stunnel tried to load it in the normal way. Please find the debug output bellow, # ./stunnel stunnel-engine.conf 2006.04.19 13:39:22 LOG7[1261:3086812864]: Enabling support for engine 'tpm' DEBUG e_tpm_err.c:295 ERR_load_TPM_strings DEBUG e_tpm_err.c:298 TPM_lib_error_code is 136 2006.04.19 13:39:22 LOG7[1261:3086812864]: Initializing engine DEBUG e_tpm.c:336 tpm_engine_init LOG_DEBUG TSPI ../tcsd_api/clntside.c:58 Sending TSP packet to host localhost. LOG_DEBUG TSPI ../tcsd_api/clntside.c:74 Connecting to 127.0.0.1 LOG_DEBUG TSPI ../tcsd_api/tcstp.c:390 TCS_OpenContext_RPC_TP: Received TCS Context: 0xa0ef791d 2006.04.19 13:39:22 LOG7[1261:3086812864]: Engine initialized 2006.04.19 13:39:22 LOG7[1261:3086812864]: Engine closed 2006.04.19 13:39:22 LOG7[1261:3086812864]: Snagged 64 random bytes from /root/.rnd DEBUG e_tpm.c:1151 tpm_rand_bytes getting 1024 bytes LOG_DEBUG TSPI ../tcsd_api/tcstp.c:2488 TCSP_GetRandom_TP: TCS Context: 0xa0ef791d 2006.04.19 13:39:23 LOG7[1261:3086812864]: Wrote 1024 new random bytes to /root/.rnd DEBUG e_tpm.c:1171 tpm_rand_status 2006.04.19 13:39:23 LOG7[1261:3086812864]: RAND_status claims sufficient entropy for the PRNG 2006.04.19 13:39:23 LOG6[1261:3086812864]: PRNG seeded successfully DEBUG e_tpm.c:736 tpm_rsa_init 2006.04.19 13:39:23 LOG7[1261:3086812864]: Certificate: /root/dk/CVS/070406/applications/openssl_tpm_engine/TpmKey.crt 2006.04.19 13:39:23 LOG7[1261:3086812864]: Key file: /root/dk/CVS/070406/applications/openssl_tpm_engine/TpmKey.key 2006.04.19 13:39:23 LOG3[1261:3086812864]: error stack: 140B3009 : error:140B3009:SSL routines:SSL_CTX_use_RSAPrivateKey_file:PEM lib 2006.04.19 13:39:23 LOG3[1261:3086812864]: SSL_CTX_use_RSAPrivateKey_file: 906D06C: error:0906D06C:PEM routines:PEM_read_bio:no start line It is obvious from the debug that the Stunnel expect a private key file in PEM format but the whole issue is that the private key I have (created using TPM) is not in a PEM format, it is an encrypted file with root key used to encrypt stays inside the TPM hardware. I suppose we need to provide a key load function which loads the key into the TPM rather than using Stunnel. It will be a great help if someone could provide me with some pointers on how to solve this, or please let me know if I am missing out something. Also do ask me if you need any further clarifications. Many thanks, Dinesh Kallath, CISSP Research Professional, Security Research Centre BT Group Chief Technology Office ___________________________ Tel : +44 (0) 1473 643476 Fax : +44 (0) 1473 646886 Mob: +44 (0) 7952144553 Email: dinesh.kallath(a)bt.com Post : PP:2A, B28, Adastral Park, Ipswich IP5 3RE. British Telecommunications plc Registered office: 81 Newgate Street London EC1A 7AJ Registered in England no. 1800000 This electronic message contains information from British Telecommunications plc which may be privileged or confidential. The information is intended to be for the use of the individual(s) or entity named above. If you are not the intended recipient be aware that any disclosure, copying, distribution or use of the contents of this information is prohibited. If you have received this electronic message in error, please notify us by telephone or email (to the numbers or address above) immediately. Activity and use of the British Telecommunications plc email system is monitored to secure its effective operation and for other lawful business purposes. Communications using this system will also be monitored and may be recorded to secure effective operation and for other lawful business purposes.

1 0

stunnel-4.15 doesn't work on Mac OS X 10.4.6
by SUENAGA Hiroki 18 Apr '06

18 Apr '06

Dear all, I found a problem that the stunnel-4.15 doesn't work on the Mac OS X 10.4.6. I think this is a BUG. My environment is: - OS: Mac OS X 10.4.6/Intel - Compiler: gcc 4.0.1 (from XCode 2.2) - OpenSSL: 0.9.7i (Mac OS X original version) - Usage: connect to the submission port. The detail of BUG is the following: In protocol.c::smtp_client() line 138, fdputline() called with argument line = "EHLO localhost." This is a string literal. 130 static void smtp_client(CLI *c) { ... 138 fdputline(c, c->remote_fd.fd, "EHLO localhost"); ... In network.c::fdputline() line 565, the literal is passed to 1st argument of safeconcat(). 556 void fdputline(CLI *c, int fd, char *line) { .... 565 safeconcat(line, crlf); ... In common.h line 311, safeconcat() is #define'ed as: 311 #define safeconcat(dst, src) \ 312 (dst[STRLEN-1]='\0', strncat((dst), (src), \ STRLEN-strlen(dst)-1)) This means the macro safeconcat() write to string literal. This is the BUG. In Mac OS X envoronment, string literal is NOT writable. So protection fault occurs at this code. On the other environment, this breaks somewhere in the heap or in the stack. This may also cause a crash. In stunnel-4.14, fdprintf() is used instead of fdputline(). fdprint() have own writable buffers on stack, and the string literal is copied to the stack. This works fine. The work around is using own writable buffers in fdputline() like fdprint() in stunnel-4.14. But it seems to be a performance tuning why fdputline() is used. So I don't know what is the correct way to fix this BUG.... I hope this BUG is fixed in the future release. If this was already reported, please ignore this mail. Thank you. Rgards, ---------- Internet Initiative Japan Inc. Research & Development Section, Product Development Division, SEIL Business Unit SUENAGA Hiroki <hsuenaga(a)iij.ad.jp>

2 2

STunnel and the MAC
by Bryan Bearden 15 Apr '06

15 Apr '06

I am working on setting up a mac server and having trouble getting the stunnel to work on the mac. I installed Darwin Ports and am running it under there. It appears to be running and my stunnel.conf file appears to be correct, but I cannot get it to use stunnel. Any ideas? I'm trying to connect to MySQL 4. thanks... Bryan

1 0

Choosing local 'source' IP? (again)
by Don Werve 14 Apr '06

14 Apr '06

This is a repost, but the response I got last time said, 'use the "local" option'...which does not appear to work. Ok, so here's the problem: I've got stunnel handling SSL for a web-based service that is otherwise too stupid to be able to talk https. On the same piece of hardware that runs this service (Virtuozzo if you must know), I've set up stunnel to listen to ETH0_IP:443 and wrapper them to ETH0_IP:80, which is where VZ listens. This works wonderfully. Unfortunately, I need to set up a redirection rule inside this service that bounces people back to the https:// page if they try to come in over http. Since said service isn't managing its own SSL, I have no solid way of determining which connections are wrappered by stunnel and which connections come in over the wire, save for by the IP address of the connection. Now, each piece of iron has two ethernet adapters, so I got the bright idea to try and get stunnel to 'appear' to be connecting from the IP address of the second adapter. So, I set 'local = ETH1_IP' in stunnel.conf, and...nada. Still shows as connecting from ETH0_IP. I've screwed with socket options, read the source code for Stunnel, and so on. Having read the source code and refreshed my knowledge of socket() and connect(), is it really even possible to do what I want if stunnel is running on the same system? Oh, a ptrace of stunnel also yields the following: socket(PF_INET, SOCK_STREAM, IPPROTO_IP) = 8 fcntl64(8, F_GETFL) = 0x2 (flags O_RDWR) fcntl64(8, F_SETFL, O_RDWR|O_NONBLOCK) = 0 bind(8, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("[ETH1_IP]")}, 16) = 0 connect(8, {sa_family=AF_INET, sin_port=htons(80), sin_addr=inet_addr("[ETH0_IP]")}, 16) = -1 EINPROGRESS (Operation now in progress) ...and the connection inside of $CRAP_SYSTEM shows the connection as coming from ETH0_IP. So, I'm kind of at my wits' end here. Any possible way to make this work? Thanks-in-advance! -- Don Werve <donw(a)iradeon.net> Chief Systems Administrator / Systems Architect

1 0

stunnel4 stealing my ports (listening on ports it shouldn't be)
by Eric Brown 12 Apr '06

12 Apr '06

I have two stunnel processes. One listing on 443 -> localhost:80. The other listening on 8000 -> localhost:7001. (See configs below.) However, when I stop the process listening on port 7001, both stunnel processes start listening on 7001 (and a few other ports). I'm completely confused as to why this would happen. This is what I see normally: mta4-tiny-support (epgate)[6]$ sudo lsof -p 25253 -p 25255 -a -i4 Password: COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME stunnel4 25253 root 6u IPv4 3439450 TCP *:8000 (LISTEN) stunnel4 25255 root 6u IPv4 3439456 TCP *:https (LISTEN) This is what I see when stunnel appears to go bizerk: (after stopping the process that was listening on ports *:7000, localhost:7001 & localhost:7005) COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME stunnel4 22788 root 5u IPv4 1269640 TCP 127.0.0.1:4341- >127.0.0.1:19996 (CLOSE_WAIT) stunnel4 22788 root 6u IPv4 1273504 TCP 127.0.0.1:7001 (LISTEN) stunnel4 22788 root 7u IPv4 1273476 TCP 127.0.0.1:4606- >127.0.0.1:19995 (ESTABLISHED) stunnel4 22788 root 9u IPv4 1273477 TCP 127.0.0.1:19995- >127.0.0.1:4606 (ESTABLISHED) stunnel4 22788 root 11u IPv4 1273507 TCP *:7000 (LISTEN) stunnel4 22788 root 17u IPv4 1273515 TCP 127.0.0.1:7005 (LISTEN) stunnel4 22788 root 19u IPv4 1274895 TCP 127.0.0.1:7001- >127.0.0.1:1793 (CLOSE_WAIT) stunnel4 22788 root 24u IPv4 1275089 TCP *:8000 (LISTEN) stunnel4 22790 root 5u IPv4 1269640 TCP 127.0.0.1:4341- >127.0.0.1:19996 (CLOSE_WAIT) stunnel4 22790 root 6u IPv4 1273504 TCP 127.0.0.1:7001 (LISTEN) stunnel4 22790 root 7u IPv4 1273476 TCP 127.0.0.1:4606- >127.0.0.1:19995 (ESTABLISHED) stunnel4 22790 root 9u IPv4 1273477 TCP 127.0.0.1:19995- >127.0.0.1:4606 (ESTABLISHED) stunnel4 22790 root 11u IPv4 1273507 TCP *:7000 (LISTEN) stunnel4 22790 root 17u IPv4 1273515 TCP 127.0.0.1:7005 (LISTEN) stunnel4 22790 root 19u IPv4 1274895 TCP 127.0.0.1:7001- >127.0.0.1:1793 (CLOSE_WAIT) stunnel4 22790 root 24u IPv4 1275094 TCP *:443 (LISTEN) Active Internet connections (only servers) Proto Recv-Q Send-Q Local Address Foreign Address State User Inode PID/Program name tcp 0 0 127.0.0.1:7008 0.0.0.0:* LISTEN 0 1281907 24848/python tcp 0 0 0.0.0.0:8000 0.0.0.0:* LISTEN 0 1275089 22788/stunnel4 tcp 0 0 0.0.0.0:1025 0.0.0.0:* LISTEN 0 1681 - tcp 0 0 0.0.0.0:963 0.0.0.0:* LISTEN 0 1556 1205/rpc.statd tcp 0 0 0.0.0.0:111 0.0.0.0:* LISTEN 0 1238 986/portmap tcp 0 0 0.0.0.0:80 0.0.0.0:* LISTEN 0 1281911 24848/python tcp 0 0 127.0.0.1:7025 0.0.0.0:* LISTEN 0 1281130 24605/master tcp 0 0 127.0.0.1:7026 0.0.0.0:* LISTEN 0 1281542 24677/python tcp 0 0 0.0.0.0:818 0.0.0.0:* LISTEN 0 1356 1061/ypbind tcp 0 0 127.0.0.1:7027 0.0.0.0:* LISTEN 0 1281731 24777/python tcp 0 0 127.0.0.1:7028 0.0.0.0:* LISTEN 0 1281910 24848/python tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN 0 1486 1195/sshd tcp 0 0 0.0.0.0:7000 0.0.0.0:* LISTEN 0 1273507 22788/stunnel4 tcp 0 0 127.0.0.1:5432 0.0.0.0:* LISTEN 400 1269115 20559/postmaster tcp 0 0 172.16.80.213:25 0.0.0.0:* LISTEN 0 1281126 24605/master tcp 0 0 172.16.80.212:25 0.0.0.0:* LISTEN 0 1280994 24584/master tcp 0 0 127.0.0.1:7001 0.0.0.0:* LISTEN 0 1273504 22788/stunnel4 tcp 0 0 0.0.0.0:443 0.0.0.0:* LISTEN 0 1275094 22790/stunnel4 tcp 0 0 127.0.0.1:955 0.0.0.0:* LISTEN 0 1518 1201/famd tcp 0 0 127.0.0.1:7005 0.0.0.0:* LISTEN 0 1273515 22788/stunnel4 tcp 0 0 127.0.0.1:7006 0.0.0.0:* LISTEN 0 1281540 24677/python tcp 0 0 127.0.0.1:7007 0.0.0.0:* LISTEN 0 1281724 24777/python I'm running stunnel 4.09 on Debian 3.1/Linux 2.6 Kernel. My configurations are: pid=/var/run/epgate_admin_stunnel4.pid cert=/epgate/etc/ssl/admin.stunnel [Admin] accept=8000 connect=7001 and: pid=/var/run/epgate_quarantine_stunnel4.pid cert=/epgate/etc/ssl/quar.stunnel [Quarantine] accept=443 connect=80

1 0

Deamon signal on close connection
by Ondra Medek 05 Apr '06

05 Apr '06

Hi, I'm using stunnel4 in daemon mode. It runs a bash script "service" for each connection. I want to stop the script, when the client closes the connection. Does the stunnel4 send any signal to child process when the connection is closed? My script receives no signal. Or is it a problem of bash? The only way how can the script find, that connection is broken is by reading the standart input, but I don't want to read any standart input. Thanks P.S. a snip of stunnel4 config: [aswconsole] accept = 1002 exec = /home/ondrejm/SeptimOS/ServerConsole/aswc-new/opt/asw/console/service

1 0

Having problems with an special configuration
by Michael Spiegle 04 Apr '06

04 Apr '06

Hi all, I've got a strange thing (at least I think so) that i'm trying to do with stunnel and it doesn't work. Here's the layout... Feel free to skip the next 3 paragraphs to get to the root of the problem: I'm using stunnel to provide SSL for a set of webservers behind a pair of LVS loadbalancers using keepalived. lets say I have a public IP of 10.0.0.1. I have an iptables rule which marks all 443 traffic to this IP with a fwmark, then keepalived dumps the traffic to the stunnel after seeing this fwmark. I have configured LVS to use direct-routing so that the stunnel server can see the original public IP which was requested (so it can serve up the right cert). Now, here is where things break. Internally, lets say stun is setup with an IP 192.168.0.10, and the LVS server has an interface on this VLAN with the IP 192.168.0.5. In my keepalived configuration for the correct fwmark, I dump the traffic to stun on 192.168.0.10. When I test this by creating a connection to the public IP address (10.0.0.1), LVS dumps the traffic to stunnel as planned and stunnel sees something like the following in TCPDUMP: Websurfer's_IP:123131 ----> 10.0.0.1:443 Stunnel refuses to do anything with the traffic. I assumed this was becuase the IP 10.0.0.1 wasn't configured anywhere on the stunnel box itself. I assigned this IP to the loopback interface as an alias, and it still doesn't work. I've straced/debug-logged stunnel and it doesn't do anything - almost like it doesn't think it should service the traffic. If I run netstat - plan, stunnel is listening to 0.0.0.0:443. So, why doesn't stunnel want to do anything with this traffic? uname -a: Linux ssl-server.domain.com 2.4.29-1.2smp #1 SMP Fri Mar 18 15:45:07 EST 2005 i686 unknown Running redhat 7.3. Glibc: glibc-2.2.5-44 glibc-common-2.2.5-44 openssl version: 0.9.6b configuration: cert = /nfs/ssl/pem/default chroot = /nfs certdir = /ssl/pem hashdir = yes pid = /run/stunnel.pid setuid = nobody setgid = nogroup output = /var/log/stunnel.log [https] accept = 443 connect = *:80 httpprotocol = yes TIMEOUTclose = 1 IP on loopback: bash-2.05a# ip addr list 1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 brd 127.255.255.255 scope host lo inet 10.0.0.1/32 brd 10.0.0.255 scope global lo:0 ... ... ... Michael Spiegle mike(a)nauticaltech.com 310.686.4570

1 0

(stunnel-4.14) poor performance with openssl engine
by Zhuang Yuyao 03 Apr '06

03 Apr '06

Hi, I've just added openssl engine support for Cavium Nitrox lite PCI card. Here are the speed test output: [engine] stands for speed test output with engine [soft] stands for speed test output without engine (openssl internal software implementation used) ~ # openssl speed -evp [rc4|aes256|sha1] [-engine /root/libcavium.so] ...... The 'numbers' are in 1000s of bytes per second processed. type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes rc4[engine] 6339.20k 25376.00k 100966.40k 3757056.00k 919961.60k rc4[soft] 6056.66k 7017.34k 7228.37k 7339.13k 7356.14k aes-256-cbc[engine] 3178.40k 25465.60k 50636.80k 194201.60k 2088140.80k aes-256-cbc[soft] 2430.36k 2572.93k 2597.45k 2617.27k 2609.82k sha1[engine] 774.20k 3115.20k 16486.40k 27033.60k 285593.60k sha1[soft] 915.74k 2622.71k 5678.93k 7942.23k 9041.75k ~ # openssl speed rsa [-engine /root/libcavium.so] sign verify sign/s verify/s rsa[soft] 512 bits 0.013810s 0.001349s 72.4 741.2 rsa[soft] 1024 bits 0.079200s 0.004342s 12.6 230.3 sign verify sign/s verify/s rsa[engine] 512 bits 0.000037s 0.000022s 26725.8 45181.5 rsa[engine] 1024 bits 0.000046s 0.000043s 21866.7 23489.1 The speed test shows than using engine is much much faster than software implementations. but while testing with stunnel, the result is really poor. I am still using apache benchmark 2 as the test tool. #ab2 -c 50 -n 100 https://192.168.10.5/test_1k.html (a html file whose size is 1Kbytes) #ab2 -c 50 -n 100 https://192.168.10.5/test_4k.html (a html file whose size is 4Kbytes) #ab2 -c 50 -n 100 https://192.168.10.5/test_50k.html (a html file whose size is 50Kbytes) #ab2 -c 50 -n 100 https://192.168.10.5/test_100k.html (a html file whose size is 100Kbytes) #ab2 -c 50 -n 100 https://192.168.10.5/test_200k.html (a html file whose size is 200Kbytes) Here is the result: Algorithm engine 1K 4K 50K 100K 200K AES-256-CBC+SHA1 no 14.12 13.21 7.92 4.48 2.17 (Requests per second) AES-256-CBC+SHA1 yes 51.10 23.34 7.30 3.12 1.91 (Requests per second) the data says everything. while using engine with stunnel, we got some performance increasement for processing small files, but processing bigger files is even slower than software implementation (it's quite strange, because proccessing large data blocks with hardware engine is assumed to be much faster than proccessing small data blocks). I did the same test for several times, but the result is the same. something I can be sure: 1) while doing stunnel engine test, all of the algorithms are processed by the engine. I can see it by debugging information printed on console screen. 2) the client running ab2 is a AMD64 with 1G RAM, and the web server (apache) used in the test above is an XEON with 2G RAM. so the bottleneck should be neither ab2 nor apache. So here are my questions: Has anyone tried stunnel with a hardware acceleartor? How about the performance? BTW, attached to this mail is a little patch for stunnel-4.14 which allows the engine directive in stunnel.conf to be set to either a engine id or a shared object path refer to the engine (ie. engine = /root/libcavium.so) Best regards, Zhuang Yuyao --- stunnel-4.14/src/ssl.c Thu Oct 27 09:42:54 2005 UTC +++ stunnel/src/ssl.c Tue Mar 28 08:48:03 2006 UTC @@ -122,6 +143,21 @@ } #if (SSLEAY_VERSION_NUMBER >= 0x00907000L) && defined(HAVE_OSSL_ENGINE_H) +static ENGINE *load_dynamic_engine(const char *engine) +{ + ENGINE *e = ENGINE_by_id("dynamic"); + if (e) + { + if (!ENGINE_ctrl_cmd_string(e, "SO_PATH", engine, 0) + || !ENGINE_ctrl_cmd_string(e, "LOAD", NULL, 0)) + { + ENGINE_free(e); + e = NULL; + } + } + return e; +} + static void init_engine(void) { ENGINE *e; @@ -133,9 +169,12 @@ } e=ENGINE_by_id(options.engine); if(!e) { + e = load_dynamic_engine(options.engine); + if (!e) { s_log(LOG_ERR, "Engine not found"); exit(1); } + } if(!ENGINE_init(e)) { s_log(LOG_ERR, "Engine can't be initialized"); ENGINE_free(e);

2 1

autostart stunnel at boot
by christof mestdag 03 Apr '06

03 Apr '06

hi is there another solution for booting stunnel without writting a custom rc5 script ? DO i have to make changes to make it works as a daemon? -> stunnel.conf client=yes output=/var/log/stunnel.log [ldaps] accept=localhost:389 connect=192.168.1.101:636

1 0