Amanda 3.3.8 and 3.3.9 packages for SUSE 12 are FUBAR
Installing the packages for Amanda 3.3.8 and/or 3.3.9 depends on a /usr/local/bin/perl. I didn't proceed with installing 3.3.8, but 3.3.9 then has some show-stoppers once installed on a SLE 12. As in the first two directives in the first script in the 15 minute setup are *BROKEN*, in a critical way. I'm guessing all the others are as well. I don't want to manually edit every file in the amanda official packages. First two lines: 1: #!/usr/local/bin/perl 23: use lib '/root/localperl/lib/site_perl/5.16.3'; The rest of the include lines don't work, as nobody has localperl version 5.16.3 by default. Might I suggest adding a user zmanda on the OBS platform, and having it build the suse versions. It'll reduce maintenance costs for whomever manages those builds now, and also ensure that new versions will be available, and will build and install for the vast majority of SUSE users.
Forum is not sending new user mails
I attempted to register on the amanda forums at forums.zmanda.com, and have been waiting a solid 5-ish minutes now and no e-mail. A minute delay is normal. 5 minute is worrisome. Somebody should poke the configuration and see what's wrong.
Re: tapes and external disks: how to combine all this
Use USB3, UASP, or esata for better performance over just USB, especially for spinning rust.It might be preferred to get two or more desktop drives, and a hard drive "toaster" (They come in 1, 2, and 4 slice varieties) to pop them into to simplify changing the drives around on the weekly. Also make sure before or after you swap the disks you force a level 0 for each DLE on the backup. On Thu, Jul 14, 2016 at 7:40 PM, Jon LaBadie wrote: > On Thu, Jul 14, 2016 at 07:12:38PM +0200, Stefan G. Weichinger wrote: >> Am 2016-07-13 um 22:19 schrieb Jon LaBadie: >> >> > I've not used it on Linux, but an Overlay FS (OFS) was introduced >> > in the kernel about 2 years ago. With that, your two drives >> > could have their own fixed mount points. Then their root dirs >> > could be overlayed so both trees appear under a single directory. >> > If one drive was not there, only half the vtapes would appear. >> >> I have to research if it is possible to have both layers writeable. >> >> I like the udev-rules-way better. >> Have the 2 disks in fstab: >> > Ahh, I finally remembered how I did it when I had vtapes on > 2 external usb disks plus part of one internal hd (not the > holding disk drive). > > Each was mounted wherever, but in the same place each time. > What amanda used was a single "changer" dir of "slots" that > were symbolic links to the vtape in that slot on its > corresponding disk. This too did not change. > > I think I had about 80 slots (and vtapes), 30 ea on usb and > 20 on hd. It did not have to be, but the slot numbering > also matched my vtape numbering. If a usb drive was not > mounted it just seemed to amanda that the slots were empty. > > If the usb drive were mounted, but went to sleep due to > inactivity, I was pleasently surprised that nothing bad > happened. What ever call woke up the disk held and waited > for the disk to be available, about 10-15 sec. > > Jon > -- > Jon H. LaBadie j...@jgcomp.com > 11226 South Shore Rd. (703) 787-0688 (H) > Reston, VA 20190 (703) 935-6720 (C)
RE: Recovering from test backup gives error: amidxtaped: critical (fatal): Insecure dependency in unlink while running setgid at /usr/local/share/perl/5.18.2/Amanda/Changer/disk.pm line 480.
So I don't know what fixed it, but I set up ssh recovery, tried to recover, and succeeded... So I switched back to local, and succeeded again.. Problem seems to be resolved for now. Never found a root cause. On Mar 23, 2015 8:04 AM, "Joi L. Ellis" wrote: > How to identify setgid/setuid stuff: > > > > root@Macropus:/# ls -la /usr/sbin/am* /usr/lib/amanda/* | grep rws > > -rwsr-xr-- 1 root backup 18880 Jan 7 2014 /usr/lib/amanda/calcsize > > -rwsr-xr-- 1 root backup 48288 Jan 7 2014 /usr/lib/amanda/dumper > > -rwsr-xr-- 1 root backup 10528 Jan 7 2014 /usr/lib/amanda/killpgrp > > -rwsr-xr-- 1 root backup 60536 Jan 7 2014 /usr/lib/amanda/planner > > -rwsr-xr-- 1 root backup 10536 Jan 7 2014 /usr/lib/amanda/rundump > > -rwsr-xr-- 1 root backup 10584 Jan 7 2014 /usr/lib/amanda/runtar > > -rwsr-xr-- 1 root backup 60720 Jan 7 2014 /usr/sbin/amcheck > > > > Look for the ‘s’ in the permissions. That denotes a file which is setuid > or setgid, either of which will trigger Taint mode in a perl script. Most > of the above are compiled binaries, but as they are running setuid, any > perl scripts they call will also be setuid and thus perl’s taint mode will > be active automatically, regardless of –T being in the hash-bang line or > not. > > > > Line 480 mentioned in your traceback has to do with deleting a directory > in the virtual changer structure on disk. I haven’t peered at the code to > see where the $drive variable is picking up a taint flag. > > > > You can find more about this with ‘man perlsec’, there is good, if brief, > discussion of taint mode and input de-tainting in that page, plus links to > more detailed tutorials. > > > > I’m running Amanda 3.3.3 on Ubuntu 14.04.02 and I’m not having any of > these issues with perl v5.18.2. > > > > *From:* owner-amanda-us...@amanda.org [mailto: > owner-amanda-us...@amanda.org] *On Behalf Of *Schlacta, Christ > *Sent:* Friday, March 20, 2015 19:54 > *Cc:* amanda-users > *Subject:* Re: Recovering from test backup gives error: amidxtaped: > critical (fatal): Insecure dependency in unlink while running setgid at > /usr/local/share/perl/5.18.2/Amanda/Changer/disk.pm line 480. > > > > I never set anything setgid. Something must be set setgid by Amanda during > the install. Any ideas how to figure out what's setgid? > > On Mar 20, 2015 4:41 PM, "Jean-Louis Martineau" > wrote: > > What is setgid? > > There should be nothing setgid when running amrecover/amidxtaped. > > Jean-Louis > > On 03/20/2015 01:10 PM, Schlacta, Christ wrote: > > So Running an amrecover after following the 15 minute guide gives me a > strange error, and I'm unable to recover.. > > > > summary: > > Add files to backup using amrecover, run extract. > > extract prompts me to load vtape, which I click yes.. loading vtape > results in: > > > > amrecover> extract > > > > Extracting files using tape drive changer on host localhost. > > The following tapes are needed: DailySet1-6 > > > > Extracting files using tape drive changer on host localhost. > > Load tape DailySet1-6 now > > Continue [?/Y/n/s/d]? > > Got no header and data from server, check in amidxtaped.*.debug and > amandad.*.debug files on server > > amrecover> exit > > > > > > After this, I check amandad.*.debug, and find absolutely nothing.. but in > amidxtaped.*.debug, I get the following content: > > > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: pid 24896 ruid 63998 > euid 63998 version 3.3.6: start at Fri Mar 20 20:02:49 2015 > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << > FEATURES=9efefbff3f > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << > CONFIG=DailySet1 > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << > LABEL=DailySet1-6:1 > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << FSF=1 > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << HEADER > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << DEVICE=changer > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << HOST=^www01$ > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << DISK=^/var/www$ > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << > DATESTAMP=20150320091506 > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << END > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: pid 24896 ruid 63998 > euid 63998 version 3.3.6: rename at Fri Mar 20 20:02:49 2015 > > Fri Mar 20
Re: Recovering from test backup gives error: amidxtaped: critical (fatal): Insecure dependency in unlink while running setgid at /usr/local/share/perl/5.18.2/Amanda/Changer/disk.pm line 480.
I never set anything setgid. Something must be set setgid by Amanda during the install. Any ideas how to figure out what's setgid? On Mar 20, 2015 4:41 PM, "Jean-Louis Martineau" wrote: > What is setgid? > > There should be nothing setgid when running amrecover/amidxtaped. > > Jean-Louis > > On 03/20/2015 01:10 PM, Schlacta, Christ wrote: > > So Running an amrecover after following the 15 minute guide gives me a > strange error, and I'm unable to recover.. > > summary: > Add files to backup using amrecover, run extract. > extract prompts me to load vtape, which I click yes.. loading vtape > results in: > > amrecover> extract > > Extracting files using tape drive changer on host localhost. > The following tapes are needed: DailySet1-6 > > Extracting files using tape drive changer on host localhost. > Load tape DailySet1-6 now > Continue [?/Y/n/s/d]? > Got no header and data from server, check in amidxtaped.*.debug and > amandad.*.debug files on server > amrecover> exit > > > After this, I check amandad.*.debug, and find absolutely nothing.. but > in amidxtaped.*.debug, I get the following content: > > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: pid 24896 ruid > 63998 euid 63998 version 3.3.6: start at Fri Mar 20 20:02:49 2015 > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << > FEATURES=9efefbff3f > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << > CONFIG=DailySet1 > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << > LABEL=DailySet1-6:1 > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << FSF=1 > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << HEADER > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << DEVICE=changer > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << HOST=^www01$ > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << DISK=^/var/www$ > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << > DATESTAMP=20150320091506 > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << END > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: pid 24896 ruid 63998 > euid 63998 version 3.3.6: rename at Fri Mar 20 20:02:49 2015 > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: chg-disk: Dir > /var/backups/vtape/DailySet1/ > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: chg-disk: Using > statefile '/var/backups/vtape/DailySet1//state' > Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: critical (fatal): > Insecure dependency in unlink while running setgid at > /usr/local/share/perl/5.18.2/Amanda/Changer/disk.pm line 480. > > amidxtaped: Insecure dependency in unlink while running setgid at > /usr/local/share/perl/5.18.2/Amanda/Changer/disk.pm line 480. > > /usr/lib/amanda/libamanda-3.3.6.so(+0x305e6)[0x7f78382945e6] > /lib/x86_64-linux-gnu/libglib-2.0.so.0(g_logv+0x1b1)[0x7f7837fabae1] > /lib/x86_64-linux-gnu/libglib-2.0.so.0(g_log+0x82)[0x7f7837fabd72] > > /usr/local/share/perl/5.18.2/auto/Amanda/MainLoop/libMainLoop.so(+0x56d5)[0x7f783208b6d5] > /lib/x86_64-linux-gnu/libglib-2.0.so.0(+0x49703)[0x7f7837fa5703] > > /lib/x86_64-linux-gnu/libglib-2.0.so.0(g_main_context_dispatch+0x135)[0x7f7837fa4ce5] > /lib/x86_64-linux-gnu/libglib-2.0.so.0(+0x49048)[0x7f7837fa5048] > > /lib/x86_64-linux-gnu/libglib-2.0.so.0(g_main_loop_run+0x6a)[0x7f7837fa530a] > > /usr/local/share/perl/5.18.2/auto/Amanda/MainLoop/libMainLoop.so(run_c+0x11)[0x7f783208a5cd] > > /usr/local/share/perl/5.18.2/auto/Amanda/MainLoop/libMainLoop.so(_wrap_run_c+0xf9)[0x7f783208ce75] > /usr/lib/libperl.so.5.18(Perl_pp_entersub+0x596)[0x7f7839710866] > /usr/lib/libperl.so.5.18(Perl_runops_standard+0x16)[0x7f7839708e86] > /usr/lib/libperl.so.5.18(perl_run+0x384)[0x7f78396a1844] > /usr/bin/perl(main+0x149)[0x400dd9] > /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf5)[0x7f78392b3ec5] > /usr/bin/perl[0x400e11] > > > I tried downgrading from 3.3.7 to 3.3.6, with no joy. This is on an > almost perfectly stock ubuntu 14.04 install. The only variation from the > quick start guide is that I've configured ssh authentication for > amandabackup and local authentication for amrecover > > >
Recovering from test backup gives error: amidxtaped: critical (fatal): Insecure dependency in unlink while running setgid at /usr/local/share/perl/5.18.2/Amanda/Changer/disk.pm line 480.
So Running an amrecover after following the 15 minute guide gives me a strange error, and I'm unable to recover.. summary: Add files to backup using amrecover, run extract. extract prompts me to load vtape, which I click yes.. loading vtape results in: amrecover> extract Extracting files using tape drive changer on host localhost. The following tapes are needed: DailySet1-6 Extracting files using tape drive changer on host localhost. Load tape DailySet1-6 now Continue [?/Y/n/s/d]? Got no header and data from server, check in amidxtaped.*.debug and amandad.*.debug files on server amrecover> exit After this, I check amandad.*.debug, and find absolutely nothing.. but in amidxtaped.*.debug, I get the following content: Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: pid 24896 ruid 63998 euid 63998 version 3.3.6: start at Fri Mar 20 20:02:49 2015 Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << FEATURES=9efefbff3f Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << CONFIG=DailySet1 Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << LABEL=DailySet1-6:1 Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << FSF=1 Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << HEADER Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << DEVICE=changer Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << HOST=^www01$ Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << DISK=^/var/www$ Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << DATESTAMP=20150320091506 Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: CTL << END Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: pid 24896 ruid 63998 euid 63998 version 3.3.6: rename at Fri Mar 20 20:02:49 2015 Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: chg-disk: Dir /var/backups/vtape/DailySet1/ Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: chg-disk: Using statefile '/var/backups/vtape/DailySet1//state' Fri Mar 20 20:02:49 2015: thd-0x19abc00: amidxtaped: critical (fatal): Insecure dependency in unlink while running setgid at /usr/local/share/perl/5.18.2/Amanda/Changer/disk.pm line 480. amidxtaped: Insecure dependency in unlink while running setgid at /usr/local/share/perl/5.18.2/Amanda/Changer/disk.pm line 480. /usr/lib/amanda/libamanda-3.3.6.so(+0x305e6)[0x7f78382945e6] /lib/x86_64-linux-gnu/libglib-2.0.so.0(g_logv+0x1b1)[0x7f7837fabae1] /lib/x86_64-linux-gnu/libglib-2.0.so.0(g_log+0x82)[0x7f7837fabd72] /usr/local/share/perl/5.18.2/auto/Amanda/MainLoop/libMainLoop.so(+0x56d5)[0x7f783208b6d5] /lib/x86_64-linux-gnu/libglib-2.0.so.0(+0x49703)[0x7f7837fa5703] /lib/x86_64-linux-gnu/libglib-2.0.so.0(g_main_context_dispatch+0x135)[0x7f7837fa4ce5] /lib/x86_64-linux-gnu/libglib-2.0.so.0(+0x49048)[0x7f7837fa5048] /lib/x86_64-linux-gnu/libglib-2.0.so.0(g_main_loop_run+0x6a)[0x7f7837fa530a] /usr/local/share/perl/5.18.2/auto/Amanda/MainLoop/libMainLoop.so(run_c+0x11)[0x7f783208a5cd] /usr/local/share/perl/5.18.2/auto/Amanda/MainLoop/libMainLoop.so(_wrap_run_c+0xf9)[0x7f783208ce75] /usr/lib/libperl.so.5.18(Perl_pp_entersub+0x596)[0x7f7839710866] /usr/lib/libperl.so.5.18(Perl_runops_standard+0x16)[0x7f7839708e86] /usr/lib/libperl.so.5.18(perl_run+0x384)[0x7f78396a1844] /usr/bin/perl(main+0x149)[0x400dd9] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf5)[0x7f78392b3ec5] /usr/bin/perl[0x400e11] I tried downgrading from 3.3.7 to 3.3.6, with no joy. This is on an almost perfectly stock ubuntu 14.04 install. The only variation from the quick start guide is that I've configured ssh authentication for amandabackup and local authentication for amrecover
amgpgcrypt and gpg agent/smart cards plus multiple key pairs
I've been reading up on amgpgcrypt. There's surprisingly little, and surprisingly less current documentation on this little extra. I want to use amgpgcrypt, but I want to get it set up 1) so that backups are encrypted to multiple key pairs, and 2) so that the keys can be stored on smart cards and issued to backup operators on an as-needed basis. Unfortunately, I can't find any documentation on whether this setup is workable or not. So my question is this: Is this a workable solution? Should it be viable for a remote administrator to SSH in (forwarding their GPG agent, of course), and run some restore commands, being prompted at their local terminal for their smart card PIN, and successfully complete the restore, or is the script written in such a way as to make this an unworkable solution?
Re: pre/post scripting
Or you could, you know, add Amanda to the sudoers.d folder On Jul 9, 2014 11:21 AM, "Stefan G. Weichinger" wrote: > Am 09.07.2014 16:17, schrieb Stefan G. Weichinger: > > > > Would anyone mind sharing some real world scripts he uses with amanda? > > > > I think of stopping/starting DBs or something like that. > > When I think about this I assume that some sudo-trickery might be > needed, right? > > The amanda-user won't be allowed to stop/start mysql, for example, so I > think I this will need using sudo or maybe adding the amanda-user to > some group with the needed permissions. > > How do you solve this? > > I will do some tests tomorrow and have a look. > > Stefan > >
MySQL Backups?
I'm looking at backing up MySQL alongside some other files. I know it used to be possible to make Amanda back up mysql databases right alongside other files, but I can only find the plugin for postgresql now, and not for MySQL. All I'm finding is a separate, non-integrated solution to back up mysql separately. That's all well and good for local backups, but I'm trying to throw them into amazon S3 together. What am I missing?
Amanda integrity and erc
What mechanisms does amanda provide to guarantee the following: 1) what was on the original drive is what is sent to the backup server 2) what was sent to the backup server is what's written to the archive 3) what's written to the archive is what's rerouted from backup 4) that any errors encountered at any of the above, or any other step in the backup process can be recovered from, at least up to a certain level of resiliency I've scoured the faq and manuals and don't recall seeing answers to any of this.
7zip instead of tar?
Is it possible, or even better, supported to use .7z instead of .tar for the backup archive format? I know .7z gets incredible compression rates and provides the archive layer functionality as well, and includes an encryption mechanism as well.
