http://permalink.gmane.org/gmane.comp.web.aolserver hourly 1 1901-01-01T00:00+00:00 http://gmane.org/img/gmane-25t.png http://gmane.org http://permalink.gmane.org/gmane.comp.web.aolserver/16542 <pre>Is your aolserver installation in a system-wide place? It's in /usr/local/aolserver451 which is not a system-wide known place... and as I like to maintain and build the web stack myself, I tend to like this approach. Having said this, both (in standard place versus anywhere else) can work if you use some of the things I'll describe now. The libs created against aolserver and tcl are dynamic libs, so ld.so is used to load and use them. /etc/ld.so.conf should list the places that the machine owner feels should be considered the standard places. So, if /usr/local/aolserver451/lib is in this file, it's considered a system-wide place, and libs in that dir will be found without further ado. If it's not, the best way is not to alter /etc/ld.so.conf, but to add the lib dir to the env var LD_LIBRARY_PATH. If you do this before trying to run nsd, the libs in /usr/local/aolserver451 will be pulled in exactly as if that dir were in /etc/ld.so.conf. -Jim On 5/13/12, Klaus Hofeditz ]project-open[ &lt;klaus.hofeditz&lt; at &gt;proje</pre> Jim 2012-05-14T06:57:37 http://permalink.gmane.org/gmane.comp.web.aolserver/16541 <pre>Your aolserver is looking for nssha1.so in this specific location: /usr/local/aolserver451/bin/nssha1.so Check your folder directory if the file exists. Make sure that the libnssha1.so is in /usr/local/aolserver451/lib too. On Monday, May 14, 2012 5:22:00 AM UTC+8, Klaus Hofeditz ]project-open[ wrote: ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/_______________________________________________ aolserver-talk mailing list aolserver-talk&lt; at &gt;lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/aolserver-talk </pre> Sep Ng 2012-05-13T22:32:39 http://permalink.gmane.org/gmane.comp.web.aolserver/16540 <pre>Hi all, I am trying to install AOLSERVER 4.5.1 on Ubuntu 12.04 (LTS). While nslog loads ok, nssha1 fails: [13/May/2012:17:09:35][15372.3073791680][-main-] Notice: modload: loading '/usr/local/aolserver451/bin/nssha1.so' [13/May/2012:17:09:35][15372.3073791680][-main-] Warning: modload: could not find Ns_ModuleInit in /usr/local/aolserver451/bin/nssha1.so [13/May/2012:17:09:35][15372.3073791680][-main-] Fatal: modload: failed to load module '/usr/local/aolserver451/bin/nssha1.so' Any ideas where to start digging? Tx for your support! Klaus root&lt; at &gt;abc:/usr/local/src/aolserver-4.5.1/nssha1# make install NSHOME=/usr/local/aolserver451 gcc -O2 -Wall -Wno-implicit-int -fPIC -pipe -I/usr/local/aolserver451/include -I/usr/local/aolserver451/include -DNO_CONST -DPACKAGE_NAME=\"tcl\" -DPACKAGE_TARNAME=\"tcl\" -DPACKAGE_VERSION=\"8.5\" -DPACKAGE_STRING=\"tcl\ 8.5\" -DPACKAGE_BUGREPORT=\"\" -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_S</pre> Klaus Hofeditz ]project-open[ 2012-05-13T21:22:00 http://permalink.gmane.org/gmane.comp.web.aolserver/16539 <pre>Thank you for your explanation Jeff. I will continue to look for where the handle was released. It must have happened on the same thread, yes? The same thread that handled the request, yes? On Saturday, May 5, 2012 12:58:40 AM UTC+8, Jeff Rogers wrote: ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/_______________________________________________ aolserver-talk mailing list aolserver-talk&lt; at &gt;lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/aolserver-talk </pre> Sep Ng 2012-05-05T00:11:29 http://permalink.gmane.org/gmane.comp.web.aolserver/16538 <pre> This means that there is no handle associated with the key "nsdb0". I'd guess this is because something called [ns_db releasehandle] when you weren't expecting it, but it's also possible that the handle was passed to a different interpreter (e.g., through a nsv) rather than acquired with [ns_db gethandle], which won't work. It's unrelated. "max connections exceeded" happens when a conn thread has served as many requests as it is configured to do and exits. -J ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ aolserver-talk mailing list aolserver-talk&lt; at &gt;lists.sourceforge.net https://lists.sourceforge.net/lists/</pre> Jeff Rogers 2012-05-04T16:58:40 http://permalink.gmane.org/gmane.comp.web.aolserver/16537 <pre>I'll just go ahead and post the solution to this problem. Apparently there's a library conflict between my Oracle install and OpenSSL. I solved it by setting LD_PRELOAD=/usr/lib/libcrypto.so.0.9.8 before running aolserver. On Tuesday, May 1, 2012 11:16:24 AM UTC+8, Sep Ng wrote: ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/_______________________________________________ aolserver-talk mailing list aolserver-talk&lt; at &gt;lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/aolserver-talk </pre> Sep Ng 2012-05-01T20:59:53 http://permalink.gmane.org/gmane.comp.web.aolserver/16536 <pre>Every once in a while on my aolserver logs, I find this error and I'm not sure what it means. Typically, I see it when performing a database operation and then aolserver drops into an error like: invalid database_id: nsdb0 Is this related to another log entry: max connections exceeded? Is the database handle not valid? Hoping someone can shed some light on the nature of this error. Thanks! ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/_______________________________________________ aolserver-talk mailing list aolserver-talk&lt; at &gt;lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/aolserver-talk </pre> Sep Ng 2012-04-30T05:28:55 http://permalink.gmane.org/gmane.comp.web.aolserver/16535 <pre>Hi, I've been looking at one aolserver install which seems to be crashing when I run with SSL certificates. The weird thing is that if I run it as root, the startup goes fine, which leads me to believe it is possibly a permission issue. The crash happens right when nsd attempts to generate 512-bit keys. Has this happened to anybody before? Thanks! ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/_______________________________________________ aolserver-talk mailing list aolserver-talk&lt; at &gt;lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/aolserver-talk </pre> Sep Ng 2012-05-01T03:16:24 http://permalink.gmane.org/gmane.comp.web.aolserver/16534 <pre>The default maximum length for a single line in the header is 4096 bytes, which your cookie will end up slightly longer than. An error message should be logged in the server log, something line "max line exceeded". It looks like the code just closes the socket when that happens, tho I have a recollection of someone recently adding in patches to make that cleaner and return an http error, but I'm not certain of that. You can increase the max line length in the server config, in the driver config section: ns_section ns/server/YOUR_SERVER/module/nssock ns_param maxline 8192 -J David Osborne wrote: ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ ___________</pre> Jeff Rogers 2012-04-26T16:05:49 http://permalink.gmane.org/gmane.comp.web.aolserver/16533 <pre>sorry.. prematurely sent the email... as I was saying... the patch described there would be useful for you, if tls is causing the crashes. It can be that the tls code you have is not thread safe. Best, On Thu, Apr 26, 2012 at 3:44 PM, Victor Guerra &lt;vguerra&lt; at &gt;wu.ac.at&gt; wrote: </pre> Victor Guerra 2012-04-26T13:46:17 http://permalink.gmane.org/gmane.comp.web.aolserver/16532 <pre>Dear Wolfgang, Take a look at this aolserver conversation here: http://www.mail-archive.com/aolserver&lt; at &gt;listserv.aol.com/msg12356.html I guess the patch described there would On Thu, Apr 26, 2012 at 3:10 PM, Wolfgang Winkler &lt; wolfgang.winkler&lt; at &gt;digital-concepts.com&gt; wrote: </pre> Victor Guerra 2012-04-26T13:44:17 http://permalink.gmane.org/gmane.comp.web.aolserver/16531 <pre>Hi! We have various high volume servers with AOLserver installations. Most of them are running stable but in all of our latest installs we have problems with the stability, especially on startup. I've recompiled all sources on a clean (tcl free) installation with: * tcl 8.5.11 (and 8.5.9 on another machine) * postgres 9.0.1 (and 9.1.3 on another machine) * tls 1.6 * thread 2.6.6 * aolserver 4.5.1 (src package and latest github version) I've checked all loaded libraries with strace and there is only one tcl version on each of the boxes. When I'm starting nsd, I get the following backtrace for 3 out of 10 times with the github version, the package version just states " received fatal signal 11" or "alloc: invalid block": [18/Apr/2012:10:48:28][25051.1675499280][-nssock:driver-] Notice: nssock: listening on 91.118.87.98:8000 *** glibc detected *** /usr/local/aolserver/bin/nsd: double free or corruption (fasttop): 0x0000000001203760 *** ======= Backtrace: ========= /lib/libc.so.6(+0x71ad6)[0x7fa79a</pre> Wolfgang Winkler 2012-04-26T13:10:07 http://permalink.gmane.org/gmane.comp.web.aolserver/16530 <pre>Hi! We have various high volume servers with AOLserver installations. Most of them are running stable but in all of our latest installs we have problems with the stability, especially on startup. I've recompiled all sources on a clean (tcl free) installation with: * tcl 8.5.11 (and 8.5.9 on another machine) * postgres 9.0.1 (and 9.1.3 on another machine) * tls 1.6 * thread 2.6.6 * aolserver 4.5.1 (src package and latest github version) I've checked all loaded libraries with strace and there is only one tcl version on each of the boxes. When I'm starting nsd, I get the following backtrace for 3 out of 10 times with the github version, the package version just states " received fatal signal 11" or "alloc: invalid block": [18/Apr/2012:10:48:28][25051.1675499280][-nssock:driver-] Notice: nssock: listening on 91.118.87.98:8000 *** glibc detected *** /usr/local/aolserver/bin/nsd: double free or corruption (fasttop): 0x0000000001203760 *** ======= Backtrace: ========= /lib/libc.so.6(+0x71ad6)[0x7fa79a</pre> Wolfgang Winkler 2012-04-18T10:08:54 http://permalink.gmane.org/gmane.comp.web.aolserver/16529 <pre>Hi, Can I check if the following is an intended behaviour? We run Aolserver 4.5.1 on Debian Squeeze. The behaviour we are seeing is, when a large cookie is sent in a request, the nsd daemon doesn't seem to reply at all. (This also occurs with a large number of smaller cookies. ~50) I constructed a test case by creating a cookie file containing 1 cookie (as attached) which is very long (probably illegally long) and used wget as follows: wget --server-response --load-cookies cookies_test_single2.txt http://www.domain.co.uk:8001/tcl/search.tcl?search=test Then wget reports "HTTP request sent, awaiting response... No data received." and retries. If I remove a couple of characters from the cookie, it will work fine. (If anyone wanted to try it then obviously the domain in the cookie file would need to be changed to match yours.) This first manifested itself with the nsd daemon running behind nginx. nginx would return a 502 with the error "upstream prematurely closed connection while reading response heade</pre> David Osborne 2012-04-26T11:49:14 http://permalink.gmane.org/gmane.comp.web.aolserver/16528 <pre>--- nsd/lisp.c | 31 ++++++++++++++----------------- 1 files changed, 14 insertions(+), 17 deletions(-) diff --git a/nsd/lisp.c b/nsd/lisp.c index d67db74..b019763 100644 --- a/nsd/lisp.c +++ b/nsd/lisp.c &lt; at &gt;&lt; at &gt; -58,16 +58,16 &lt; at &gt;&lt; at &gt; static const char *RCSID = "&lt; at &gt;(#) $Header: /Users/dossy/Desktop/cvs/aolserver/nsd Ns_List * Ns_ListNconc(Ns_List *l1Ptr, Ns_List *l2Ptr) { - Ns_List *lPtr; - - if (l1Ptr != NULL) { - for (lPtr = l1Ptr; ((lPtr-&gt;rest) != NULL); lPtr = lPtr-&gt;rest) - ; - lPtr-&gt;rest = l2Ptr; - return (l1Ptr); - } else { - return (l2Ptr); - } + if (l1Ptr == NULL) +return l2Ptr; + + Ns_List *lPtr = l1Ptr; + + for (lPtr = l1Ptr; ((lPtr-&gt;rest) != NULL); lPtr = lPtr-&gt;rest) + ; + + lPtr-&gt;rest = l2Ptr; + return (l1Ptr); } &lt; at &gt;&lt; at &gt; -154,13 +154,10 &lt; at &gt;&lt; at &gt; Ns_ListNreverse(Ns_List *lPtr) Ns_List * Ns_ListLast(Ns_List *lPtr) { - if (lPtr == NULL) { - return NULL; - } else { - for (; lPtr-&gt;rest != NULL; lPtr = lPtr-&gt;rest) - ; - </pre> Ayan George 2012-04-04T22:52:10 http://permalink.gmane.org/gmane.comp.web.aolserver/16527 <pre>Sorry -- I screwed up the previous diff. I blame vi. --- nsd/lisp.c | 31 ++++++++++++++----------------- 1 files changed, 14 insertions(+), 17 deletions(-) diff --git a/nsd/lisp.c b/nsd/lisp.c index d67db74..b019763 100644 --- a/nsd/lisp.c +++ b/nsd/lisp.c &lt; at &gt;&lt; at &gt; -58,16 +58,16 &lt; at &gt;&lt; at &gt; static const char *RCSID = "&lt; at &gt;(#) $Header: /Users/dossy/Desktop/cvs/aolserver/nsd Ns_List * Ns_ListNconc(Ns_List *l1Ptr, Ns_List *l2Ptr) { - Ns_List *lPtr; - - if (l1Ptr != NULL) { - for (lPtr = l1Ptr; ((lPtr-&gt;rest) != NULL); lPtr = lPtr-&gt;rest) - ; - lPtr-&gt;rest = l2Ptr; - return (l1Ptr); - } else { - return (l2Ptr); - } + if (l1Ptr == NULL) +return l2Ptr; + + Ns_List *lPtr = l1Ptr; + + for (lPtr = l1Ptr; ((lPtr-&gt;rest) != NULL); lPtr = lPtr-&gt;rest) + ; + + lPtr-&gt;rest = l2Ptr; + return (l1Ptr); } &lt; at &gt;&lt; at &gt; -154,13 +154,10 &lt; at &gt;&lt; at &gt; Ns_ListNreverse(Ns_List *lPtr) Ns_List * Ns_ListLast(Ns_List *lPtr) { - if (lPtr == NULL) { - return NULL; - } else { - for</pre> Ayan George 2012-04-04T22:54:00 http://permalink.gmane.org/gmane.comp.web.aolserver/16526 <pre>Hi, Thanks for the reply. How did you solve the problem. I have had to remove php from the nsd file. Thanks, Thorpe On Mar 16, 2012, at 1:28 PM, wkliang wrote: ------------------------------------------------------------------------------ This SF email is sponsosred by: Try Windows Azure free for 90 days Click Here http://p.sf.net/sfu/sfd2d-msazure _______________________________________________ aolserver-talk mailing list aolserver-talk&lt; at &gt;lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/aolserver-talk </pre> Thorpe Mayes 2012-03-16T18:57:04 http://permalink.gmane.org/gmane.comp.web.aolserver/16525 <pre>Hello, the same issue had been encountered by me. guess that was caused by php+mysql while doing some clean up stuff. regards, wkliang &lt; at &gt; taipei. On Thu, Mar 15, 2012 at 10:12 PM, Thorpe Mayes &lt;tmayes&lt; at &gt;ecognizant.net&gt; wrote: ------------------------------------------------------------------------------ This SF email is sponsosred by: Try Windows Azure free for 90 days Click Here http://p.sf.net/sfu/sfd2d-msazure _______________________________________________ aolserver-talk mailing list aolserver-talk&lt; at &gt;lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/aolserver-talk </pre> wkliang 2012-03-16T18:28:45 http://permalink.gmane.org/gmane.comp.web.aolserver/16524 <pre> Hi, I reinstalled php5 (I did not remove anything before reinstalling) with the curl extension enabled. After doing so, I got this message upon shutting down the server: [15/Mar/2012:08:29:55][13455.3083933376][-main-] Notice: driver: stopped: nssock *** glibc detected *** /usr/local/aolserver/bin/nsd: corrupted double-linked list: 0x09e5ec38 *** ======= Backtrace: ========= /lib/libc.so.6[0xb7d76894] /lib/libc.so.6(__libc_free+0x78)[0xb7d79f6f] /usr/local/aolserver/bin/libphp5.so(tsrm_shutdown+0x9f)[0xb6e90a9f] /usr/local/aolserver/bin/libphp5.so[0xb6f5b92c] /usr/local/aolserver/lib/libnsd.so[0xb7f542b9] /usr/local/aolserver/lib/libnsd.so[0xb7f54410] /usr/local/aolserver/lib/libnsthread.so(NsThreadMain+0x2f)[0xb7f3707f] /usr/local/aolserver/lib/libnsthread.so[0xb7f38384] /lib/libpthread.so.0[0xb7e7c433] /lib/libc.so.6(__clone+0x5e)[0xb7ddba1e] ======= Memory map: ======== 08048000-08049000 r-xp 00000000 08:02 4178137 /usr/local/aolserver/bin/nsd 08049000-0804a000 rwxp 00000000 08:02 4178137 /usr/</pre> Thorpe Mayes 2012-03-15T14:12:39 http://permalink.gmane.org/gmane.comp.web.aolserver/16523 <pre> The case for windows is special, since the crash happens there during the unload of the dll's, which is highly platform specific. The situation on unixes is different. In our configurations (mostly linux with 32bit and 64 bit, rhel/fedora/ubuntu, as well some mac os x), the shutdown works fine since many years (although we do not call it normally from the control port). It is not the case, that Tcl_Finalize() in recent Tcl versions is inherently broken and "needs to be fixed". It is the situation as Jim points out, that some modules/packages might register handlers that have some bugs, or - which would be my primary suspect - some loaded module/package has a bug (overwritten some memory, double frees, ...) which manifests during cleanup. Cleanup is highly sensitive to bugs in the memory management. The source of the problem should be fixed, not the symptom. It is not unlikely that the same bug will hit you in some other cases as well... -gustaf -------------------------------------------</pre> Gustaf Neumann 2012-03-06T13:00:08 http://permalink.gmane.org/gmane.comp.web.aolserver/16522 <pre> Howdy, The more I think about it, the config option makes good sense as a compromise. Gustaf's point about a crash hinting at other problems is a strong reason to leave it working in some fashion instead of just dumping altogether as I had in the past. /* BEGIN OPINION ... consider enjoying Facebook instead of reading this babble... */ But, I will re-iterate one thing: The manner in which Tcl handles cleanup makes it technically impossible to avoid all possible crashes assuming you allow Tcl to be used in other programs of which you do not have control of all the code. To be clear, the pthread model that Aolserver follows (i.e., iterative calls to cleanups) cannot be guaranteed to work either -- you could conceive of a condition where thread cleanup A re-initalizes B which re-initalizes A, etc. That's why there's a retry count, set to 5. It attempts to catch the normal case of going through the cleanups just once, tries again in the off-chance something got re-initalized, and once again in the ve</pre> Jim Davidson 2012-03-05T17:11:45 Search the mailing list at Gmane query http://search.gmane.org/?group=$group=gmane.comp.web.aolserver