I've made a number of changes to fix issues in 2.1.0. Can some people please try git head? See git.freeradius.org for how to get a "tar" file. If there are no complaints, we can issue 2.1.1 tomorrow. The new command tool "radmin" is starting to get insane. As in "insanely useful". Alan DeKok.
Alan DeKok wrote:
I've made a number of changes to fix issues in 2.1.0. Can some people please try git head? See git.freeradius.org for how to get a "tar" file.
If there are no complaints, we can issue 2.1.1 tomorrow.
The new command tool "radmin" is starting to get insane. As in "insanely useful".
Alan DeKok. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
It certainly builds on our RPM buildslave for RHEL5 x86_64. I'll fire it up and run our tests after lunch. Any chance we can get bug 572 in 2.1.1? I've just sent a new version of the patch for git HEAD
Hi,
I've made a number of changes to fix issues in 2.1.0. Can some people please try git head? See git.freeradius.org for how to get a "tar" file.
PS - it appears thatt he .tar.gz file thats available to download from that location is for a 2.0.6 release. just thought I'd mention it.... alan
A.L.M.Buxey@lboro.ac.uk wrote:
PS - it appears thatt he .tar.gz file thats available to download from that location is for a 2.0.6 release. just thought I'd mention it....
The tar file on the web site is old. The web site has pointers to downloading updated tar files from github. Alan DeKok.
hi, just a quick note for those folk out there running lots of PERL DBI things etc with their FreeRADIUS systems on CentOS/RHEL - theres been a problem with PERL with this distro for some time - massive speed penalty - over 100x slower than ubuntu/suse/etc etc this quick PERL script should complete in a sub-second #!/usr/bin/perl use overload q(<) => sub {}; my %h; for (my $i=0; $i<50000; $i++) { $h{$i} = bless [ ] => 'main'; print STDERR '.' if $i % 1000 == 0; } on an affected box it takes several seconds....anyway, just to let people know that I can now confirm CentOS have release their patch/fix so all is back to normal. cannot confirm RHEL. anyway, all PERL DBI stuff is now back to the full speed. alan
Thanks for info.... #yum update perl is a solution On Thu, Sep 18, 2008 at 9:20 AM, <A.L.M.Buxey@lboro.ac.uk> wrote:
hi,
just a quick note for those folk out there running lots of PERL DBI things etc with their FreeRADIUS systems on CentOS/RHEL - theres been a problem with PERL with this distro for some time - massive speed penalty - over 100x slower than ubuntu/suse/etc etc
this quick PERL script should complete in a sub-second
#!/usr/bin/perl use overload q(<) => sub {}; my %h; for (my $i=0; $i<50000; $i++) { $h{$i} = bless [ ] => 'main'; print STDERR '.' if $i % 1000 == 0; }
on an affected box it takes several seconds....anyway, just to let people know that I can now confirm CentOS have release their patch/fix so all is back to normal. cannot confirm RHEL. anyway, all PERL DBI stuff is now back to the full speed.
alan - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
Hi, just tried a GIT checkout and got something nasty. config files work fine with the 2.1.0 release *** glibc detected *** radiusd: double free or corruption (!prev): 0x083dea60 *** ======= Backtrace: ========= /lib/libc.so.6[0x386b16] /lib/libc.so.6(cfree+0x90)[0x38a070] /usr/lib/libfreeradius-radius-2.1.1.so[0x5695f2] /usr/lib/libfreeradius-radius-2.1.1.so(rbtree_free+0x32)[0x569642] radiusd(realms_free+0x52)[0x806e852] radiusd(realms_init+0x554)[0x806ee24] radiusd(read_mainconfig+0x6aa)[0x805b69a] radiusd(main+0x37e)[0x806067e] /lib/libc.so.6(__libc_start_main+0xdc)[0x333dec] radiusd[0x804d861] ======= Memory map: ======== 00110000-00119000 r-xp 00000000 fd:00 6225969 /lib/libcrypt-2.5.so 00119000-0011a000 r-xp 00008000 fd:00 6225969 /lib/libcrypt-2.5.so 0011a000-0011b000 rwxp 00009000 fd:00 6225969 /lib/libcrypt-2.5.so 0011b000-00142000 rwxp 0011b000 00:00 0 00142000-0014b000 r-xp 00000000 fd:00 655409 /lib/libnss_files-2.5.so 0014b000-0014c000 r-xp 00008000 fd:00 655409 /lib/libnss_files-2.5.so 0014c000-0014d000 rwxp 00009000 fd:00 655409 /lib/libnss_files-2.5.so 00300000-0031a000 r-xp 00000000 fd:00 655451 /lib/ld-2.5.so 0031a000-0031b000 r-xp 00019000 fd:00 655451 /lib/ld-2.5.so 0031b000-0031c000 rwxp 0001a000 fd:00 655451 /lib/ld-2.5.so 0031e000-0045b000 r-xp 00000000 fd:00 655457 /lib/libc-2.5.so 0045b000-0045d000 r-xp 0013c000 fd:00 655457 /lib/libc-2.5.so 0045d000-0045e000 rwxp 0013e000 fd:00 655457 /lib/libc-2.5.so 0045e000-00461000 rwxp 0045e000 00:00 0 00463000-00465000 r-xp 00000000 fd:00 6225959 /lib/libdl-2.5.so 00465000-00466000 r-xp 00001000 fd:00 6225959 /lib/libdl-2.5.so 00466000-00467000 rwxp 00002000 fd:00 6225959 /lib/libdl-2.5.so 00469000-0047c000 r-xp 00000000 fd:00 6225960 /lib/libpthread-2.5.so 0047c000-0047d000 r-xp 00012000 fd:00 6225960 /lib/libpthread-2.5.so 0047d000-0047e000 rwxp 00013000 fd:00 6225960 /lib/libpthread-2.5.so 0047e000-00480000 rwxp 0047e000 00:00 0 004ab000-004c1000 r-xp 00000000 fd:00 6225965 /lib/libselinux.so.1 004c1000-004c3000 rwxp 00015000 fd:00 6225965 /lib/libselinux.so.1 004c5000-00500000 r-xp 00000000 fd:00 6225964 /lib/libsepol.so.1 00500000-00501000 rwxp 0003a000 fd:00 6225964 /lib/libsepol.so.1 00501000-0050b000 rwxp 00501000 00:00 0 0050d000-0051f000 r-xp 00000000 fd:00 12288567 /usr/lib/libz.so.1.2.3 0051f000-00520000 rwxp 00011000 fd:00 12288567 /usr/lib/libz.so.1.2.3 00522000-00535000 r-xp 00000000 fd:00 6225957 /lib/libnsl-2.5.so 00535000-00536000 r-xp 00012000 fd:00 6225957 /lib/libnsl-2.5.so 00536000-00537000 rwxp 00013000 fd:00 6225957 /lib/libnsl-2.5.so 00537000-00539000 rwxp 00537000 00:00 0 0053b000-00541000 r-xp 00000000 fd:00 3708318 /usr/lib/libltdl.so.3.1.4 00541000-00542000 rwxp 00005000 fd:00 3708318 /usr/lib/libltdl.so.3.1.4 00559000-00575000 r-xp 00000000 fd:00 6848547 /usr/lib/libfreeradius-radius-2.1.1.so 00575000-00576000 rwxp 0001c000 fd:00 6848547 /usr/lib/libfreeradius-radius-2.1.1.so 00576000-00577000 rwxp 00576000 00:00 0 0057a000-00697000 r-xp 00000000 fd:00 6225967 /lib/libcrypto.so.0.9.8b 00697000-006aa000 rwxp 0011c000 fd:00 6225967 /lib/libcrypto.so.0.9.8b 006aa000-006ad000 rwxp 006aa000 00:00 0 006af000-006be000 r-xp 00000000 fd:00 6225963 /lib/libresolv-2.5.so 006be000-006bf000 r-xp 0000e000 fd:00 6225963 /lib/libresolv-2.5.so 006bf000-006c0000 rwxp 0000f000 fd:00 6225963 /lib/libresolv-2.5.so 006c0000-006c2000 rwxp 006c0000 00:00 0 006c4000-006c6000 r-xp 00000000 fd:00 6225966 /lib/libcom_err.so.2.1 006c6000-006c7000 rwxp 00001000 fd:00 6225966 /lib/libcom_err.so.2.1 006c9000-006d4000 r-xp 00000000 fd:00 6225956 /lib/libgcc_s-4.1.2-20080102.so.1 006d4000-006d5000 rwxp 0000a000 fd:00 6225956 /lib/libgcc_s-4.1.2-20080102.so.1 00769000-0078e000 r-xp 00000000 fd:00 3704091 /usr/lib/libk5crypto.so.3.1 0078e000-0078f000 rwxp 00025000 fd:00 3704091 /usr/lib/libk5crypto.so.3.1 00791000-00793000 r-xp 00000000 fd:00 6225962 /lib/libkeyutils-1.2.so 00793000-00794000 rwxp 00001000 fd:00 6225962 /lib/libkeyutils-1.2.so 00796000-007c3000 r-xp 00000000 fd:00 12288541 /usr/lib/libgssapi_krb5.so.2.2 007c3000-007c4000 rwxp 0002d000 fd:00 12288541 /usr/lib/libgssapi_krb5.so.2.2 007c6000-00856000 r-xp 00000000 fd:00 12288037 /usr/lib/libkrb5.so.3.3 00856000-00859000 rwxp 0008f000 fd:00 12288037 /usr/lib/libkrb5.so.3.3 0085b000-00863000 r-xp 00000000 fd:00 3704090 /usr/lib/libkrb5support.so.0.1 00863000-00864000 rwxp 00007000 fd:00 3704090 /usr/lib/libkrb5support.so.0.1 008c4000-00905000 r-xp 00000000 fd:00 6225968 /lib/libssl.so.0.9.8b 00905000-00909000 rwxp 00040000 fd:00 6225968 /lib/libssl.so.0.9.8b 00e26000-00e27000 r-xp 00e26000 00:00 0 [vdso] 08048000-0807d000 r-xp 00000000 fd:00 7077954 /usr/sbin/radiusd 0807d000-08080000 rw-p 00034000 fd:00 7077954 /usr/sbin/radiusd 082fa000-083e7000 rw-p 082fa000 00:00 0 b7e00000-b7e21000 rw-p b7e00000 00:00 0 b7e21000-b7f00000 ---p b7e21000 00:00 0 b7f79000-b7f7f000 rw-p b7f79000 00:00 0 bfa1b000-bfa3b000 rw-p bfa1b000 00:00 0 [stack] Aborted it barfs when reading a home_server entry in proxy.conf home_server test0 { ipaddr = 192.168.0.139 port = 1812 type = "auth+acct" secret = "drstrangeloveisaclassic" response_window = 30 max_outstanding = 65536 zombie_period = 40 status_check = "none" ping_interval = 30 check_interval = 30 num_answers_to_alive = 3 num_pings_to_alive = 3 revive_interval = 300 status_check_timeout = 4 } home_server_pool test { home_server = test0 } realm test.co.uk { pool = test } ..it doesnt make it as far as the home_server_pool ..... alan
A.L.M.Buxey@lboro.ac.uk wrote:
just tried a GIT checkout and got something nasty. config files work fine with the 2.1.0 release
Fixed, thanks. There was a bug in the new "list home servers" functionality. This is what it now outputs: $ ./radmin -e "show home_server list" 192.168.0.139 1812 auth alive 0 192.168.0.139 1813 acct alive 0 127.0.0.1 1812 auth alive 0 The last "0" is the number of packets that have been sent to the server that haven't (yet) received a reply. Alan DeKok.
Hi, got a problem with GIT version of 2.1.1 - it doesnt fail in debug mode (-X) but when run as a service, it soon dies (in less than 5 minutes) - as stated before, in full debug mode it runs all day. I am currently trying to discover the fault.... alan
A.L.M.Buxey@lboro.ac.uk wrote:
I am currently trying to discover the fault....
$ gdb --args radiusd -f (gdb) run ... (gdb) <arg> There were a few changes to fix problems found in 2.1.0, and a whole whack of new development to add useful new features. Unfortunately, it looks like the new development added issues of it's own... Alan DeKok.
Hi,
There were a few changes to fix problems found in 2.1.0, and a whole whack of new development to add useful new features. Unfortunately, it looks like the new development added issues of it's own...
just to report that i've got the latest GIT version now installed and running.... alan
On Wed, 2008-09-17 at 11:48 +0200, Alan DeKok wrote:
I've made a number of changes to fix issues in 2.1.0. Can some people please try git head? See git.freeradius.org for how to get a "tar" file.
The 'configure', 'make' and 'make test' all completed successfully on a CentOS 5.2 (64-bit) system with no problems. John. -- --------------------------------------------------------------- John Horne, University of Plymouth, UK Tel: +44 (0)1752 587287 E-mail: John.Horne@plymouth.ac.uk Fax: +44 (0)1752 587001
On Wed, 2008-09-17 at 11:48 +0200, Alan DeKok wrote:
I've made a number of changes to fix issues in 2.1.0. Can some people please try git head? See git.freeradius.org for how to get a "tar" file.
If there are no complaints, we can issue 2.1.1 tomorrow.
FYI - Installing 2.1.1 gives this error (but doesn't stop the installation): Creating/updating files in /usr/local/etc/raddb install: snmp.conf does not exist find: /usr/local/etc/raddb/snmp.conf: No such file or directory It seems that Makefile has 'snmp.conf' listed as a file to be installed, but no such file exists. Running 'radiusd -X' shows: Write out database with 1 new entries Data Base Updated unable to write 'random state' unable to write 'random state' MAC verified OK Not sure what it means by the 'random state' messages, or even if they are important :-) John. -- --------------------------------------------------------------- John Horne, University of Plymouth, UK Tel: +44 (0)1752 587287 E-mail: John.Horne@plymouth.ac.uk Fax: +44 (0)1752 587001
John Horne wrote:
Creating/updating files in /usr/local/etc/raddb install: snmp.conf does not exist find: /usr/local/etc/raddb/snmp.conf: No such file or directory
It seems that Makefile has 'snmp.conf' listed as a file to be installed, but no such file exists.
OK, I've fixed that.
Not sure what it means by the 'random state' messages, or even if they are important :-)
They're not important. I don't know why the complaint exists, tho. Alan DeKok.
I've made a number of changes to fix issues in 2.1.0. Can some people please try git head? See git.freeradius.org for how to get a "tar" file.
I saw that there were a few changes since this message was sent. I did a new git clone a couple hours ago and installed from there. Here is the result. As a reminder, I am using a configuration based on the robust-proxy-accounting example. FR 2.1.0 crashes for me when the remote accounting server is down. This version, however, seems to crash after a few seconds regardless of the state of the remote accounting server. OS is CentOS 5.2 (32 bit) with all updates installed. Let me know if you need any further info. Thanks, Jim L. Starting program: /usr/sbin/radiusd -f (no debugging symbols found) (no debugging symbols found) [Thread debugging using libthread_db enabled] [New Thread -1208149296 (LWP 23938)] [New Thread -1208153200 (LWP 23941)] [New Thread -1218643056 (LWP 23942)] [New Thread -1229132912 (LWP 23943)] [New Thread -1239622768 (LWP 23944)] [New Thread -1250112624 (LWP 23945)] Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -1239622768 (LWP 23944)] 0x00000000 in ?? () (gdb) thread apply all bt full Thread 6 (Thread -1250112624 (LWP 23945)): #0 0x0077b402 in __kernel_vsyscall () No symbol table info available. #1 0x0062026e in sem_wait@GLIBC_2.0 () from /lib/libpthread.so.0 No symbol table info available. #2 0x08062cbf in request_handler_thread () No symbol table info available. #3 0x0061a46b in start_thread () from /lib/libpthread.so.0 No symbol table info available. #4 0x00571dbe in clone () from /lib/libc.so.6 No symbol table info available. Thread 5 (Thread -1239622768 (LWP 23944)): #0 0x00000000 in ?? () No symbol table info available. #1 0x0806700c in debug_packet () No symbol table info available. #2 0x08068e6c in proxy_request () No symbol table info available. #3 0x0806963a in request_post_handler () No symbol table info available. #4 0x08069a18 in radius_handle_request () No symbol table info available. #5 0x08063014 in request_handler_thread () No symbol table info available. #6 0x0061a46b in start_thread () from /lib/libpthread.so.0 No symbol table info available. #7 0x00571dbe in clone () from /lib/libc.so.6 No symbol table info available. Thread 4 (Thread -1229132912 (LWP 23943)): #0 0x0077b402 in __kernel_vsyscall () No symbol table info available. #1 0x0062026e in sem_wait@GLIBC_2.0 () from /lib/libpthread.so.0 No symbol table info available. #2 0x08062cbf in request_handler_thread () No symbol table info available. #3 0x0061a46b in start_thread () from /lib/libpthread.so.0 No symbol table info available. #4 0x00571dbe in clone () from /lib/libc.so.6 No symbol table info available. Thread 3 (Thread -1218643056 (LWP 23942)): #0 0x0077b402 in __kernel_vsyscall () No symbol table info available. #1 0x0062026e in sem_wait@GLIBC_2.0 () from /lib/libpthread.so.0 No symbol table info available. #2 0x08062cbf in request_handler_thread () No symbol table info available. #3 0x0061a46b in start_thread () from /lib/libpthread.so.0 No symbol table info available. #4 0x00571dbe in clone () from /lib/libc.so.6 No symbol table info available. Thread 2 (Thread -1208153200 (LWP 23941)): #0 0x0077b402 in __kernel_vsyscall () No symbol table info available. #1 0x0062026e in sem_wait@GLIBC_2.0 () from /lib/libpthread.so.0 No symbol table info available. #2 0x08062cbf in request_handler_thread () No symbol table info available. #3 0x0061a46b in start_thread () from /lib/libpthread.so.0 No symbol table info available. #4 0x00571dbe in clone () from /lib/libc.so.6 No symbol table info available. Thread 1 (Thread -1208149296 (LWP 23938)): #0 0x0077b402 in __kernel_vsyscall () No symbol table info available. #1 0x0056abc1 in select () from /lib/libc.so.6 No symbol table info available. #2 0x0024f561 in fr_event_loop (el=0x8822368) at event.c:378 i = <value optimized out> rcode = 1 when = {tv_sec = 0, tv_usec = 750000} wake = (struct timeval *) 0xbf9ede34 read_fds = {fds_bits = {1040448, 0 <repeats 31 times>}} #3 0x080686c6 in radius_event_process () No symbol table info available. #4 0x080608f5 in main () No symbol table info available. #0 0x00000000 in ?? ()
Hi,
I saw that there were a few changes since this message was sent. I did a new git clone a couple hours ago and installed from there. Here is the result. As a reminder, I am using a configuration based on the robust-proxy-accounting example. FR 2.1.0 crashes for me when the remote accounting server is down. This version, however, seems to crash after a few seconds regardless of the state of the remote accounting server. OS is CentOS 5.2 (32 bit) with all updates installed. Let me know if you need any further info.
did you --enable-developer in the ./configure stage? all your symbols are shot out. regarding the crash - i see a similar thread issue on Fedora system. if you run the daemon in debug mode (radiusd -X) you should see it being rock solid. I think we're close to nailing it. alan
Jim L. wrote:
I saw that there were a few changes since this message was sent. I did a new git clone a couple hours ago and installed from there. Here is the result. As a reminder, I am using a configuration based on the robust-proxy-accounting example. FR 2.1.0 crashes for me when the remote accounting server is down.
I'm still not sure why. That's a bad thing to happen...
This version, however, seems to crash after a few seconds regardless of the state of the remote accounting server. OS is CentOS 5.2 (32 bit) with all updates installed. Let me know if you need any further info. ... Thread 5 (Thread -1239622768 (LWP 23944)): #0 0x00000000 in ?? () No symbol table info available. #1 0x0806700c in debug_packet ()
I've fixed that. See the recent git. Alan DeKok.
Alan, I just did a new git clone and compared it to the one I used for this test. They appear to be the same. Jim L. ----- Original Message ----- From: "Alan DeKok" <aland@deployingradius.com> To: "FreeRadius users mailing list" <freeradius-users@lists.freeradius.org> Sent: Friday, September 19, 2008 1:58 AM Subject: Re: 2.1.1-pre
Jim L. wrote:
I saw that there were a few changes since this message was sent. I did a new git clone a couple hours ago and installed from there. Here is the result. As a reminder, I am using a configuration based on the robust-proxy-accounting example. FR 2.1.0 crashes for me when the remote accounting server is down.
I'm still not sure why. That's a bad thing to happen...
This version, however, seems to crash after a few seconds regardless of the state of the remote accounting server. OS is CentOS 5.2 (32 bit) with all updates installed. Let me know if you need any further info. ... Thread 5 (Thread -1239622768 (LWP 23944)): #0 0x00000000 in ?? () No symbol table info available. #1 0x0806700c in debug_packet ()
I've fixed that. See the recent git.
Alan DeKok. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
participants (6)
-
A.L.M.Buxey＠lboro.ac.uk -
Alan DeKok -
Jim L. -
John Horne -
Marinko Tarlac -
Phil Mayers