RE: amlabel segmentation fault
Hi Jean-Louis, Sorry for the delay. I had to remove a binary install and get the source build back up again. But it all works now with your patch. Thanks! Best regards, Kervin Web Services Design, Development and Maintenance http://adevsoft.com/ -Original Message- From: Jean-Louis Martineau [mailto:martin...@zmanda.com] Sent: Friday, January 10, 2014 8:35 AM To: Kervin L. Pierre Cc: amanda-users@amanda.org Subject: Re: amlabel segmentation fault Kervin, Did you try the patch? Jean-Louis On 01/06/2014 09:13 AM, Jean-Louis Martineau wrote: Kervin, Looks like there is a bug if libcurl use GNUTLS. Can you try the attached patch? Jean-Louis On 01/05/2014 12:10 PM, Kervin L. Pierre wrote: Hi Jean-Louis, Thanks for the response. Is there a way to rebuild with '-g' for debugging support and to avoid stripping symbols? Would that help? Here is the gdb output... $ gdb --args perl /usr/local/sbin/amlabel --version GNU gdb (GDB) 7.6.1-ubuntu Copyright (C) 2013 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later http://gnu.org/licenses/gpl.html This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type show copying and show warranty for details. This GDB was configured as x86_64-linux-gnu. For bug reporting instructions, please see: http://www.gnu.org/software/gdb/bugs/... Reading symbols from /usr/bin/perl...(no debugging symbols found)...done. (gdb) run Starting program: /usr/bin/perl /usr/local/sbin/amlabel --version [Thread debugging using libthread_db enabled] Using host libthread_db library /lib/x86_64-linux-gnu/libthread_db.so.1. Program received signal SIGSEGV, Segmentation fault. 0x7489876d in ?? () from /lib/x86_64-linux-gnu/libgcrypt.so.11 (gdb) backtrace #0 0x7489876d in ?? () from /lib/x86_64-linux-gnu/libgcrypt.so.11 #1 0x748925ae in ?? () from /lib/x86_64-linux-gnu/libgcrypt.so.11 #2 0x7489004c in gcry_control () from /lib/x86_64-linux-gnu/libgcrypt.so.11 #3 0x70c75259 in device_api_init () at device.c:64 #4 0x70ebc1f2 in boot_Amanda__Device (my_perl=optimized out, cv=optimized out) at Amanda/Device.c:4676 #5 0x77b10591 in Perl_pp_entersub () from /usr/lib/libperl.so.5.14 #6 0x77b07d36 in Perl_runops_standard () from /usr/lib/libperl.so.5.14 #7 0x77aa43c0 in Perl_call_sv () from /usr/lib/libperl.so.5.14 #8 0x77aa51cc in Perl_call_list () from /usr/lib/libperl.so.5.14 #9 0x77a8faf1 in ?? () from /usr/lib/libperl.so.5.14 #10 0x77a9b7b3 in Perl_newATTRSUB () from /usr/lib/libperl.so.5.14 #11 0x77a9bfa2 in Perl_utilize () from /usr/lib/libperl.so.5.14 #12 0x77ac989e in Perl_yyparse () from /usr/lib/libperl.so.5.14 #13 0x77aa9b76 in perl_parse () from /usr/lib/libperl.so.5.14 #14 0x00400dae in main () (gdb) list 1built-in: No such file or directory. Best regards, Kervin Web Services Design, Development and Maintenance http://adevsoft.com/ -Original Message- From: Jean-Louis Martineau [mailto:martin...@zmanda.com] Sent: Sunday, January 5, 2014 9:19 AM To: Kervin L. Pierre Cc: amanda-users@amanda.org Subject: Re: amlabel segmentation fault Kevin, run amlabel under gdb: $ gdb perl (gdb) run /usr/local/sbin/amlabel --version (gdb) bt (gdb) list Jean-Louis On 01/04/2014 11:01 PM, Kervin L. Pierre wrote: Hi, This Segfault seems to occur occasionally and for different reasons. It's possible that amlabel do not handle unknown devices gracefully. But I'm still looking. My tapedevice is S3. But I just built 3.3.5 on a Ubuntu desktop and now I get a segfault whenever I run amlabel... $ perl -d /usr/local/sbin/amlabel --version Loading DB routines from perl5db.pl version 1.33 Editor support available. Enter h or `h h' for help, or `man perldebug' for more help. Signal SEGV at /usr/local/share/perl/5.14.2/Amanda/Device.pm line 12 require Amanda/Device.pm called at /usr/local/sbin/amlabel line 29 main::BEGIN() called at /usr/local/share/perl/5.14.2/Amanda/Device.pm line 0 eval {...} called at /usr/local/share/perl/5.14.2/Amanda/Device.pm line 0 Device.pm looks like # This file was automatically generated by SWIG (http://www.swig.org). # Version 2.0.4 # # Do not make changes to this file unless you know what you are doing--modify # the SWIG interface file instead. package Amanda::Device; use base qw(Exporter); use base qw(DynaLoader); require Amanda::Header; package Amanda::Devicec; bootstrap Amanda::Device; package Amanda::Device; @EXPORT = qw(); With the bootstrap Amanda line being line 12. Should that line read package Amanda::Devicec;? Any idea what the issue could be? Best regards, Kervin Web Services Design, Development and Maintenance http://adevsoft.com/
amlabel segmentation fault
Hi, This Segfault seems to occur occasionally and for different reasons. It's possible that amlabel do not handle unknown devices gracefully. But I'm still looking. My tapedevice is S3. But I just built 3.3.5 on a Ubuntu desktop and now I get a segfault whenever I run amlabel... $ perl -d /usr/local/sbin/amlabel --version Loading DB routines from perl5db.pl version 1.33 Editor support available. Enter h or `h h' for help, or `man perldebug' for more help. Signal SEGV at /usr/local/share/perl/5.14.2/Amanda/Device.pm line 12 require Amanda/Device.pm called at /usr/local/sbin/amlabel line 29 main::BEGIN() called at /usr/local/share/perl/5.14.2/Amanda/Device.pm line 0 eval {...} called at /usr/local/share/perl/5.14.2/Amanda/Device.pm line 0 Device.pm looks like # This file was automatically generated by SWIG (http://www.swig.org). # Version 2.0.4 # # Do not make changes to this file unless you know what you are doing--modify # the SWIG interface file instead. package Amanda::Device; use base qw(Exporter); use base qw(DynaLoader); require Amanda::Header; package Amanda::Devicec; bootstrap Amanda::Device; package Amanda::Device; @EXPORT = qw(); With the bootstrap Amanda line being line 12. Should that line read package Amanda::Devicec;? Any idea what the issue could be? Best regards, Kervin Web Services Design, Development and Maintenance http://adevsoft.com/
RE: aclocal fails since AMANDA_INIT_VERSION call in configure.in
Hi Dan, Thanks for looking into the issue. I'm using a fully updated Amazon Linux installation and the Autoconf package that comes with it is 2.63. Checking http://rpmfind.net/linux/rpm2html/search.php?query=autoconf It looks like most stable distributions haven't updated past 2.63 as yet either. Is there any way to keep the minimum version = 2.63 ? It would save a lot of platform maintainers from updating automake and autoconf from source before building Amanda. Best regards, Kervin Adevsoft Inc Business Software Development http://adevsoft.com/ -Original Message- From: owner-amanda-us...@amanda.org [mailto:owner-amanda- us...@amanda.org] On Behalf Of Dan Locks Sent: Wednesday, June 27, 2012 4:24 AM To: amanda-users@amanda.org Subject: Re: aclocal fails since AMANDA_INIT_VERSION call in configure.in I ran across this same error while updating amanda's packages. On 06/11/2012 09:16 AM, John Hein wrote: I see no reference to m4_divert_diversion in AMANDA_INIT_VERSION ... snip Ah... from the autoconf info pages... . . Unfortunately older versions of Automake (e.g., Automake 1.4) did not quote the names of these macros. Therefore, when `m4' finds something like `AC_DEFUN(AM_TYPE_PTRDIFF_T, ...)' in `aclocal.m4', `AM_TYPE_PTRDIFF_T' is expanded, replaced with its Autoconf definition. Fortunately Autoconf catches pre-`AC_INIT' expansions, and complains, in its own words: $ cat configure.ac AC_INIT([Example], [1.0], [bug-exam...@example.org]) AM_TYPE_PTRDIFF_T $ aclocal-1.4 $ autoconf aclocal.m4:17: error: m4_defn: undefined macro: _m4_divert_diversion aclocal.m4:17: the top level autom4te: m4 failed with exit status: 1 $ Modern versions of Automake no longer define most of these macros, and properly quote the names of the remaining macros. If you must use an old Automake, do not depend upon macros from Automake as it is simply not its job to provide macros (but the one it requires itself): . . This gives a hint that I needed to... Investigate your automake installation. Specifically, it seems that with AMANDA_INIT_VERSION, we now require autoconf =2.64. I actually tested 2.60, 2.61, 2.63, 2.64, and 2.68. The first 3 versions failed, the last two worked. I'll update configure.in to reflect the finding. Hopefully we prevent some unnecessary pain for some other source builders. Dan Locks
RE: aclocal fails since AMANDA_INIT_VERSION call in configure.in
Hello Jean-Louis, The error is from autogen. I believe the error is at least related to the new AMANDA_INIT_VERSION macro in configure.in. If I remove that first line in configure.in the then the error goes away and Amanda builds. # ./autogen See DEVELOPING for instructions on updating: * gettext macros * gnulib * libtool files ..creating file lists ..aclocal configure.in:1: error: m4_defn: undefined macro: _m4_divert_diversion configure.in:1: the top level autom4te: /usr/bin/m4 failed with exit status: 1 aclocal: autom4te failed with exit status: 1 aclocal failed # aclocal --version aclocal (GNU automake) 1.11.1 Best regards, Kervin Adevsoft Inc Business Software Development http://adevsoft.com/ -Original Message- From: Jean-Louis Martineau [mailto:martin...@zmanda.com] Sent: Monday, June 11, 2012 8:24 AM To: Kervin L. Pierre Cc: amanda-users@amanda.org Subject: Re: aclocal fails since AMANDA_INIT_VERSION call in configure.in On 06/09/2012 12:08 PM, Kervin L. Pierre wrote: I'm building on a stock Amazon Linux server with all available patches. But it seems that since the new AMANDA_INIT_VERSION macro call was added a few weeks ago to configure.in, I haven't been able to run autogen.sh without error. Removing the AMANDA_INIT_VERSION call before AC_INIT seems to be the only work around I've found. Best regards, Kervin Kevin, What error do you get? There is no autogen.sh in amanda, the program is autogen Jean-Louis
RE: what is the proper way to delete a particlur backup ( slot )
Thanks a Jean, amrmtape --erase --cleanup DailySet1 backup_data01 Did exactly what I needed. Thanks! Best regards, Kervin Adevsoft Inc Business Software Development http://adevsoft.com/ -Original Message- From: Jean-Louis Martineau [mailto:martin...@zmanda.com] Sent: Tuesday, May 01, 2012 11:56 AM To: Kervin L. Pierre Cc: amanda-users@amanda.org Subject: Re: what is the proper way to delete a particlur backup ( slot ) On 05/01/2012 11:41 AM, Kervin L. Pierre wrote: What is the proper way to delete a particular backup slot and free the space it's holding? amrmtape --erase I've tried 'amadmin DailySet1 delete ...' to remove 'slot1', but the files are still there in 'slot1'. Is it ok to 'rm -rf' in the slot folder? Are these files referenced anywhere else? I don't want to accidently corrupt the install. yes, but amanda might not know you removed them, you must also remove the label from the tapelist file. I was hoping there was a way to remove these files in a backup type independent way. Eg. I'd prefer not to have to log into AWS to delete S3 backups, etc. amrmtape --erase Jean-Louis
RE: amrecover fails with TypeError in method 'debug'...
Hi Greg, I did see your last thread, that did help thanks. Building the Alpha from Trunk, then applying the patch from... http://forums.zmanda.com/showthread.php?3691-amrecover-segmentation-fault ...fixed the 'extract' error for me. I got a stacktrace using Carp::longmess() at the point where $line is tainted in amidxtaped.pl/readline() by checking with... sub is_tainted { return ! eval { eval(# . substr(join(, @_), 0, 0)); 1 }; } This is the stack when when the line is tainted... [stacktrace] Mon Apr 30 11:39:33 2012: thd-0x10765f0: amidxtaped: pid 1305 ruid 503 euid 503 version 4.0.0alpha.svn.4682: start at Mon Apr 30 11:39:33 2012 Mon Apr 30 11:39:33 2012: thd-0x10765f0: amidxtaped: $line = '1' Mon Apr 30 11:39:33 2012: thd-0x10765f0: amidxtaped: at /usr/local/libexec/amanda/amidxtaped line 198 main::ClientService::read_command('main::ClientService=HASH(0xbf57c8)') called at /usr/local/libexec/amanda/amidxtaped line 186 main::ClientService::setup_streams('main::ClientService=HASH(0xbf57c8)') called at /usr/local/libexec/amanda/amidxtaped line 145 main::ClientService::run('main::ClientService=HASH(0xbf57c8)') called at /usr/local/libexec/amanda/amidxtaped line 1056 main::__ANON__() called at /usr/local/share/perl5/Amanda/MainLoop.pm line 819 Amanda::MainLoop::__ANON__('Amanda::MainLoop::Source=HASH(0x17e91b0)') called at /usr/local/share/perl5/Amanda/MainLoop.pm line 790 eval {...} called at /usr/local/share/perl5/Amanda/MainLoop.pm line 790 Amanda::MainLoop::run() called at /usr/local/libexec/amanda/amidxtaped line 1057 main::main() called at /usr/local/libexec/amanda/amidxtaped line 1063 Mon Apr 30 11:39:33 2012: thd-0x10765f0: amidxtaped: critical (fatal): TypeError in method 'debug', argument 1 of type 'char *' amidxtaped: TypeError in method 'debug', argument 1 of type 'char *' [/stacktrace] Best regards, Kervin Adevsoft Inc Business Software Development http://adevsoft.com/ From: Greg Copeland [mailto:gtcopel...@gmail.com] Sent: Monday, April 30, 2012 12:03 AM To: Kervin L. Pierre Cc: amanda-users@amanda.org Subject: Re: amrecover fails with TypeError in method 'debug'... On Sun, 2012-04-29 at 18:20 +, Kervin L. Pierre wrote: Hi, Look for an email from me in the archives, with subject, Re: Recovery Help Request [RESOLVED], from April 13th, 2012. For good measure, you might also want to look in the archives for a message with subject, Re: amrecover failure [RESOLVED], on April 26th, 2012. Both sets of emails, bugs, and fixes apply to 3.3.1 stable, which apparently isn't all that stable. I sincerely hope they'll push out a 3.3.2, with the patches applied for these known, show-stopper, bugs. Best of luck. Regards, I'm following the simple recovery example from the wiki... http://wiki.zmanda.com/index.php/GSWA/Recovering_Files Worked though all the issues but now I'm at the 'extract' and I'm getting an error... amrecover - amidxtaped closed the connection amrecover - can't talk to tape server: (null) Further inspection of /var/log/messages gives... amidxtaped[2784]: TypeError in method 'debug', argument 1 of type 'char *' I've found someone had run into this issue before 'June 2011'... http://archives.zmanda.com/amanda-archives/viewtopic.php?t=6929sid=f2141406dc79480edd30a1200443e9a6 They mentioned patching a Perl script but I haven't found the one they're referring to. Does anyone have instructions on how to fix or workaround this issue? I'm using 3.3.1 built from source on an update to date Amazon Linux install. Best regards, Kervin Adevsoft Inc Business Software Development http://adevsoft.com/ -- Greg Copeland g...@copelandconsulting.net Copeland Computer Consulting
Device type s3 is not known
After building from trunk to fix a few other issues, I started configuring Amazon S3 support. But amlabel reported... Device type s3 is not known Similar to this issue... https://forums.zmanda.com/archive/index.php/t-2581.html To fix this I did... make distclean yum install libcurl-devel perl-WWW-Curl ./configure --enable-s3-device make I believe part of the underlying issue is that the autoconf tests do not seem to warn about missing headers like libcurl, etc. Best regards, Kervin Adevsoft Inc Business Software Development http://adevsoft.com/
amrecover fails with TypeError in method 'debug'...
Hi, I'm following the simple recovery example from the wiki... http://wiki.zmanda.com/index.php/GSWA/Recovering_Files Worked though all the issues but now I'm at the 'extract' and I'm getting an error... amrecover - amidxtaped closed the connection amrecover - can't talk to tape server: (null) Further inspection of /var/log/messages gives... amidxtaped[2784]: TypeError in method 'debug', argument 1 of type 'char *' I've found someone had run into this issue before 'June 2011'... http://archives.zmanda.com/amanda-archives/viewtopic.php?t=6929sid=f2141406dc79480edd30a1200443e9a6 They mentioned patching a Perl script but I haven't found the one they're referring to. Does anyone have instructions on how to fix or workaround this issue? I'm using 3.3.1 built from source on an update to date Amazon Linux install. Best regards, Kervin Adevsoft Inc Business Software Development http://adevsoft.com/