mondo-devel Mailing List for Mondo Rescue

     Despite Victor Gattegno's prompt, detailed and *most* helpful reply, Somehow, I still do not have the knack for running mondoarchive on Ubuntu 11.04.  I did follow the Mondo Rescue FAQ Q#11 (see http://trac.mondorescue.org/wiki/FAQ#Q11DoesmondoworkwithDebianUbuntudistributions).  I did fix up the mindi script to use syslinux --stupid.  I do have the right versions of mondo, mindi and mindi-busybox both installed (and 'pinned' via /etc/apt/preferences).

apt-show-versions -a mondo mindi mindi-busybox
mindi 2.0.7.8-1 install ok installed
mindi 2.0.7.8-1          unknown ftp.mondorescue.org
mindi 1:2.0.4-1.1ubuntu1 natty   us.archive.ubuntu.com
No stable version
mindi/unknown uptodate 2.0.7.8-1
mindi-busybox 1.18.3-3 install ok installed
mindi-busybox 1.2.1-4  natty   us.archive.ubuntu.com
mindi-busybox 1.18.3-3 unknown ftp.mondorescue.org
No stable version
mindi-busybox/unknown uptodate 1.18.3-3
mondo 2.2.9.7-1 install ok installed
mondo 2.2.9.7-1   unknown ftp.mondorescue.org
mondo 1:2.2.7-2.1 natty   us.archive.ubuntu.com
No stable version
mondo/unknown uptodate 2.2.9.7-1


I even ssh root@localhost to ensure both real and effective user id's are *root* for:
mondoarchive -Oi -G -p `hostname`-`date +%Y%m%d` -9 -d \
/home/schulegaard/tmp/mondo -s 4480m -S /tmp -T /tmp \
-E "/mnt/usb1_1|/mnt/usb1_2|/mnt/work|/tmp|/home/schulegaard/tmp/mondo"

Which results in the usual stuff - ending like this:
...
...
Calling MINDI to create boot+data disks
Your boot loader is GRUB and it boots from /dev/sda
Waiting for 'echo hi > /tmp/mondo.tmp.15Er7D/mojo-jojo.bla.bla ; mindi   --custo
Mindi failed to create your boot+data disks.
Fatal error... Failed to generate boot+data disks
---FATALERROR--- Failed to generate boot+data disks
If you require technical support, please contact the mailing list.
See http://www.mondorescue.org for details.
The list's members can help you, if you attach that file to your e-mail.
Log file: /var/log/mondoarchive.log
Mondo has aborted.
Execution run ended; result=254
Type 'less /var/log/mondoarchive.log' to see the output log
rm: cannot remove `/tmp/mondo.tmp.15Er7D/mnt1': Device or resource busy
root@gandalf:/home/schulegaard/tmp/mondo# df
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/sda1            953178004  11277044 893482248   2% /
none                   4089988       740   4089248   1% /dev
none                   4097716      1248   4096468   1% /dev/shm
none                   4097716       244   4097472   1% /var/run
none                   4097716         0   4097716   0% /var/lock
/dev/sdc1            1922787892   3419064 1919368828   1% /mnt/usb1_1
/dev/sdc2            961465672 221931348 739534324  24% /mnt/usb1_2
/dev/sdb1            961432904    204436 961228468   1% /mnt/work
/dev/loop0              118995      1538    116188   2% /tmp/mondo.tmp.15Er7D/mnt1

Hmm.  I've attached my latest log files.  Poking around in log files makes me think that the mondoarchive mounted /dev/loop0 device is, somehow, just too small.  Is that possible?  At any rate, I manually umount /dev/loop0 after the aborted mondoarchive.  Honestly, i have no idea what role that plays for mondoarchive. 


If there's a guru out there as kind as Victor Gattegno, perhaps I'll get mondoarchive working on Ubuntu 11.04.  I must be missing something. If I might impose for just a bit more help, I sure would appreciate it. 


Cheers, Steffen


===========================

====  Victor's Response 

====  Not to shu...@ya... but from mondo-devel archive cut-n-paste
===========================

Re: [Mondo-devel] Ubuntu 11.04 mondoarchive fails & mindi .deb shows    conflicts ...From: Gattegno, Victor (GCC - Software) <victor_gattegno@hp...> - 2011-09-26 08:22

Hi Steffen, It is recommended to run mondoarchive as root, not through sudo. Don't use mondoarchive "-k FAILSAFE" option, failsafe kernel used to be provided in a separate package and is not maintained anymore. > mindi-busybox ... Version: 2.2.9.7-1
???
Latest versions on ftp://ftp.mondorescue.org/ubuntu/11.04:
mondo-2.2.9.7-1
mindi-2.0.7.8-1
mindi-busybox-1.18.3-3 About mindi versions confusion, uninstall your mindi versions, and install only mindi-2.0.7.8. About mondoarchive -E option, since mondo 2.2.9.5. there is a new syntax for -E and -I options, pipe character ('|') is used to separate the paths, instead of space character.
So you should use:
-E "/media/bf9ada79-31c0-49da-bfad-6c71f397718d|/home/schulegaard/tmp/mondo /tmp/mondo" > FATAL ERROR. Cannot unzip lib.tar.bz2
Try with mondoarchive -G option (it uses gzip instead) About ide_cd:
ide_cd is already included in IDE_MODS line, in /usr/sbin/mindi shell-script, no need to add it to FORCE_MODS. I get also "FATAL: Module ide_cd not found." with my Ubuntu 11.04, but the backup is successfull. Rgds,
Victor 

What is the current status of mindi support for Ethernet bonding?
My server has 3 three different networks each with a pair of bonded
NICs.
It would be easier to deal with mindi with my setup if I could build it
with 
support for the bonded NICs.

Alan

We have a few servers here. Two are Dell PowerEdge 700  and two are Dell 
PowerEdge 710 (rack mounted 700). Centos 5.6 64 bit has been installed 
on all four, and all four have been updated to Centos 5.7.

The latest stable mondo, mindi, buffer, afio, and mindi-busybox are 
running on all of them  The same mondo archive command was run on all of 
them to create a set of DVDs.

Booting from these DVDs created a problem which we see on all four the 
first time we try to "nuke" the system pertaining to partitioning, 
formatting where it seems to lock up. By restarting the nuke, though, 
after mondo has optimized the partition sizes, all seems well on that 
aspect.

The DVD problem we're having only on the PE 710s is that using an option 
of "-s 4480m" to create a DVD iso is not creating usable DVDs iso images.

The first one we made could not be burnt to a blank DVD on more that one 
different machines.  The copy program indicate an unsucessful burn. 
Threw that disc away.

I then changed the parm to "-s 4000m". The image could then be burned 
sucessfully, but looking at the back of the burnt DVD, which completed 
the burn, showed fragmented areas of data (concentric rings of data, not 
one large donut of data). Upon trying to nuke with this disc caused a 
non-ending restore, ending at about 70% and the same fileset each time.

I decided to try CD images, and changed the option to "-s 640m" and 
tried nuking again. This time it burned fine and nuke completed 
restoring the system.

Due to installing some other rpms later, I installed the rpmforge repos, 
and immediately got notice to update afio to something more newer than 
the rpm on the mondo site. Trying again, thinking maybe afio was the 
problem with these two machines, I created and burned new iso images. 
They seemed to burn faster and created the normal one-donut ring of data 
on the DVD. I then tried to nuke the machine, and got further, but still 
a sort of stall at some point on the DVD. Total data amounted to about 
1.5G each time.

A few things that didn't seem right and are confusing my judgement of 
what the problem really is being caused by:

Even thought the PE 700 and PE710 machines should be very close to being 
the same, I'm sure there could be varying hardware and as such different 
drivers, so the problem could be here, which is independent of anything 
Mondo does.

No matter which form of disc iso I create, CD or DVD, and whether it 
completes or not, at some time during the restore the phrase "-- MARK 
--" shows up following the "xx% to go" on the restore screen. I'm not 
sure if that's supposed to mean something like a tape mark or what.

The backup complains about not finding any of the files in 
/etc/sysconfig/network-scripts, so I have to re-address all of the NICs 
on the machine.

Keep in mind that the same job will run just fine on the PE700, creating 
perfectly normal DVD iso images.

Anyone have any clues what to look for? Sorry to be so windy this 
morning. Just hoping that someone else has seen this. No logs right now 
as I'm not sure it's really mondo. I'll review the restore logs and if I 
see anything, I'll bring this up again and attach the logs.

steve campbell

I try to carve out some time for this next week.
If I do, I will report the results.


-----Original Message-----
From: Bruno Cornec [mailto:Bru...@hp...] 
Sent: Thursday, September 29, 2011 5:31 PM
To: Mondo mailing list
Subject: Re: [Mondo-devel] Red Hat/CentOS 5.6, mindi boots, but no
functional scripts, no network

Bruno Richard said on Tue, Sep 27, 2011 at 09:17:36AM +0200:
> I had the same pb with Dell PE2950 computers and their broadcom NIC
that is not recognized 
> by mindi.

Here it seems to be the bnx2i driver which is missing in mindi's list.
Added in rev [2876]

In the mena time you can use the config file to add it (cf:
http://trac.mondorescue.org/wiki/FAQ#Q41HowdoIsupportcloningtoadifferent
system)

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           /
http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  /
http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org
http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org
http://www.medieval.org

------------------------------------------------------------------------
------
All of the data generated in your IT infrastructure is seriously
valuable.
Why? It contains a definitive record of application performance,
security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2dcopy2
_______________________________________________
Mondo-devel mailing list
Mon...@li...
https://lists.sourceforge.net/lists/listinfo/mondo-devel

Luis A. Arnaiz said on Tue, Sep 27, 2011 at 07:42:02PM +0200:

> I try to implement a solution based in Mondo. I have a machine that boot
> from a SAN disk, and I do a mondo of this disk. After, I create a new SAN
> disk and I try to recover the mondo in the new disk and boot from it, but
> all my test failed.
> 
> I would like know if it is possible with mondo and if somebody tried it and
> he did it successfully

This is possible. I had a working config with RHEL5.x and HP EVA using
multipath. You have to regenerate your initrd at restore time in order
to benefit from the new LUN references in the multipath config.

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

OK. I will report back to you asap.



Regards,


By Colomoto

El 30/09/2011, a las 01:31, Bruno Cornec <Bru...@hp...> escribió:

> Hello,
> 
> Sebastian Pereira said on Sat, Sep 17, 2011 at 12:25:25PM +0200:
> 
>> I have a problem (Huston???), the thing is that when I try to restore from a
>> NFS source the prompt for NFS disk 1 appears with the OK button, and that´s
>> it. Continuously it ask for NFS #1 and it can´t find it in the mounted NFS
>> source.
> 
> THis is  known problem with 2.2.9.7, hopefully fixed in 2.2.9.8. A beta
> is now available for you to test at
> ftp://ftp.mondorescue.org/test/rhel/5
> 
> Could you check again with that version and report back please ?
> 
> TIA,
> Bruno.
> -- 
> Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
> HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
> http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
> La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org
> 
> ------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure contains a
> definitive record of customers, application performance, security
> threats, fraudulent activity and more. Splunk takes this data and makes
> sense of it. Business sense. IT sense. Common sense.
> http://p.sf.net/sfu/splunk-d2dcopy1
> _______________________________________________
> Mondo-devel mailing list
> Mon...@li...
> https://lists.sourceforge.net/lists/listinfo/mondo-devel

Heijmans S (spir-it) said on Fri, Jul 01, 2011 at 12:34:40PM +0200:

> after booting with mondorescue.iso in expert mode, I cannot reboot normally.
> it doesn't matter before of after mondorestore.
> - exit                  gives a kernel panix
> - reboot                        does nothing
> - shutdown -ry now      restarts init processs; Welcome to init (from mindi 2.0.7.8-r2840)
>                                 and starts the mondorescue shell.
> - halt                  does indeed halt the systeem but now need to powercycle the vmware guest or server.

I'm following this in bug #499 (http://trac.mondorescue.org/ticket/499)

> Didn't have this on mondo 2.2.9.4 before, any ideas?

Not for the moment.

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

Ethan Rosenberg, PhD said on Sun, Jul 31, 2011 at 01:26:37PM -0400:

>     1] Mondo will only work for me if I declare that I am backing up to
>     a hard drive, and then give the path to the USB drive which is the
>     backup target.<br>

Are you using something like:
USB: Backup to your USB key, using gzip compression:
       mondoarchive -OU -d /dev/sda -G

>     2] Mondo fails to write an ISO file, and then does not backup any
>     data.<br>

sh: mkisofs: not found

>   </body>
> </html>

Please do not use HTML for posting on ML.

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

Alan Pittman said on Fri, Jul 29, 2011 at 08:20:21AM -0400:

> Tried running mondo archive again. Unfortunately, I got the same results. Mondoarchive doesn't seem to recognize the fact that the sshfs file system is already mounted.

Should be improved in latest beta now generating.
Cf: http://trac.mondorescue.org/ticket/496

Thanks for the logs which helped me understand where was the problem !

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

Hello Bruno !

Bruno Richard said on Thu, Sep 15, 2011 at 02:18:15PM +0200:
> I'm writting scripts to automatically save and restore Dell PE2950 with RHEL5.
> The save process is ok but I can't succeed in using mondorestore in command line.
> The following command ended without any error message but do nothing...
> [root@sirion /]# /usr/sbin/mondorestore -n "nfs://172.16.7.58:/mondosauv" -d "/altea2-v3/lv_tmpsv" -p lv_tmpsv -K 99 -I / -Z nuke
> Execution run ended; result=0
> Type 'less /var/log/mondorestore.log' to see the output log
> [root@sirion /]#
> I attached the mondorestore log file.

I'm not sure that the version you're using is able to support options at
the CLI correctly :-(

Would you mind testing with the beta of 2.2.9.8, so we could make a
point on that version, and start fixing from there (easier for me)

Cordialement,
Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

Bruno Richard said on Tue, Sep 27, 2011 at 09:17:36AM +0200:
> I had the same pb with Dell PE2950 computers and their broadcom NIC that is not recognized 
> by mindi.

Here it seems to be the bnx2i driver which is missing in mindi's list.
Added in rev [2876]

In the mena time you can use the config file to add it (cf:
http://trac.mondorescue.org/wiki/FAQ#Q41HowdoIsupportcloningtoadifferentsystem)

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

Le 30/09/2011 01:49, Bruno Cornec a écrit :
> Hello,
>
> Bruno CAPELETO said on Tue, Sep 20, 2011 at 09:43:37AM +0200:
>> Bruno Cornec a écrit :
>>> Bruno CAPELETO said on Sun, Sep 18, 2011 at 05:13:27AM +0200:
>>>
>>>> Support for SUID/SGID handling (FEATURE_SUID) [N/y/?] ->  MUST BE Y
>>> Of course, I can provide a busybox compiled with it, but I'd really like
>>> to understand why it improves stuff.
>>>
>> I didn't do anything else than answering YES in the compilation option.
>> Nothing regarding SUID bit.
>> Maybe this was not an option at all in previous versions, or this time
>> there is a bug in busybox.
>> I think it is a bug this time, because the command "id root" tells you
>> that the user root is unknown...
> I think this one is a major showstoper, and linked to the distribution
> used. I'm pretty sure we're missing a file so that the context looks
> good on Debian.
>
I don't know what I can do to help here...
>>>> 2. One needs to use authentication via ssh key and the ssh private key
>>>> of root is not copied in /.ssh (=home directory of root)
>>>> I had to copy it manually on the initrd.img ; after reboot with this new
>>>> initrd.img, the sshfs mount works fine.
>>> Line 2467 of mindi there is:
>>> 	cp -rp ~root/.ssh/*.pub ~root/.ssh/config ~root/.ssh/known* $mountpoint/.ssh 2>  /dev/null
>>>
>>> Maybe there is an issue with this line in your case. Because it should
>>> copy the pub key.
>>>
>> Yes it copies the public key, BUT for the ssh connection one needs the
>> private key ~root/.ssh/id_rsa !!!
>
> Argh :-( When I made test, I was able to give a passwd, but I guess that
> in your case, you just authenticate with the key and have no passwd
> possibility :-(
>
> This is in rev [2875] now. thanks for this feedback.
Yes : I authenticate without any password so that everything is automatic
>>>> 3. Last remark : /tmp/mondorestore.log is at the end a dead link to
>>>> /var/log/mondo-archive.log
>>> In which case ?
>>> I searched the whole code wihtout finding a ref to mondo-archive.log
>>> (except in a commented line).  I hope you don't have a mixed
>>> environment of packages and tar.gz install as it is really prone to
>>> errors.
>>>
>> I am using the package from your repository. The only thing I changed is
>> the busybox executable (that I recompiled).
> I'm at loss here. By "at the end", do you mean, end of restore before
> trying to reboot ?
>
> Bruno.
That was just before launching mondorestore : I realized that when I was 
blocked because of the issue.

Hello,

Bruno CAPELETO said on Tue, Sep 20, 2011 at 09:43:37AM +0200:
> Bruno Cornec a écrit :
> > Bruno CAPELETO said on Sun, Sep 18, 2011 at 05:13:27AM +0200:
> >
> >> Support for SUID/SGID handling (FEATURE_SUID) [N/y/?] -> MUST BE Y
> >
> > Of course, I can provide a busybox compiled with it, but I'd really like
> > to understand why it improves stuff.
> >   
> I didn't do anything else than answering YES in the compilation option.
> Nothing regarding SUID bit.
> Maybe this was not an option at all in previous versions, or this time 
> there is a bug in busybox.
> I think it is a bug this time, because the command "id root" tells you 
> that the user root is unknown...

I think this one is a major showstoper, and linked to the distribution
used. I'm pretty sure we're missing a file so that the context looks
good on Debian.

> >> 2. One needs to use authentication via ssh key and the ssh private key 
> >> of root is not copied in /.ssh (=home directory of root)
> >> I had to copy it manually on the initrd.img ; after reboot with this new 
> >> initrd.img, the sshfs mount works fine.
> >
> > Line 2467 of mindi there is:
> > 	cp -rp ~root/.ssh/*.pub ~root/.ssh/config ~root/.ssh/known* $mountpoint/.ssh 2> /dev/null
> >
> > Maybe there is an issue with this line in your case. Because it should
> > copy the pub key.
> >   
> Yes it copies the public key, BUT for the ssh connection one needs the 
> private key ~root/.ssh/id_rsa !!!


Argh :-( When I made test, I was able to give a passwd, but I guess that
in your case, you just authenticate with the key and have no passwd
possibility :-(

This is in rev [2875] now. thanks for this feedback.

> >> 3. Last remark : /tmp/mondorestore.log is at the end a dead link to 
> >> /var/log/mondo-archive.log
> >
> > In which case ? 
> > I searched the whole code wihtout finding a ref to mondo-archive.log
> > (except in a commented line).  I hope you don't have a mixed
> > environment of packages and tar.gz install as it is really prone to
> > errors.
> >   
> I am using the package from your repository. The only thing I changed is 
> the busybox executable (that I recompiled).

I'm at loss here. By "at the end", do you mean, end of restore before
trying to reboot ?

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

Hello,

Sebastian Pereira said on Sat, Sep 17, 2011 at 12:25:25PM +0200:

> I have a problem (Huston???), the thing is that when I try to restore from a
> NFS source the prompt for NFS disk 1 appears with the OK button, and that´s
> it. Continuously it ask for NFS #1 and it can´t find it in the mounted NFS
> source.

THis is  known problem with 2.2.9.7, hopefully fixed in 2.2.9.8. A beta
is now available for you to test at
ftp://ftp.mondorescue.org/test/rhel/5

Could you check again with that version and report back please ?

TIA,
Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

Jeffrey G Thomas said on Wed, Sep 28, 2011 at 04:17:16PM -0500:

> nuke denymods="aha1542,aha154x,eata"

nuke denymods="aha1542 aha154x eata" is the right syntax.

> The FAQ only tells how to deny a single module and if there is a solution, this should be updated
> http://trac.mondorescue.org/wiki/FAQ#Q36WhyismySLES10restoreblockingatfdomain

This is documented more precisely in
http://trac.mondorescue.org/browser/branches/2.2.9/mindi/README.bootparam
(Also delivered locally as part of mindi)

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

Hi Jeffrey,

Try this :
 nuke denymods="aha1542 aha154x eata"

Rgds,
Victor

-----Original Message-----
From: Jeffrey G Thomas [mailto:Jef...@ne...] 
Sent: mercredi 28 septembre 2011 23:17
To: mon...@li...
Cc: Cornec, Bruno (Open Source and Linux Technology Architect)
Subject: Re: [Mondo-devel] Debian 6 hangs at Fatal: error inserting module aha1542

> > > I've tried a few of the Debian 6 packages available but I cannot seem to get this working.
> > > FATAL: Error inserting aha1542

> > Could you send the mondoarchive.log and the mindi.log in gzip format of 
> > that error please?
Not really... the hanging freezes my machine, where would I get that log if I cannot interact?

> Yes, and try to deny that module by booting with denymods="aha1542"
Thanks for the tip; while that 
nuke denymods="aha1542"
works to deny that module, I get hung at the next module, aha152x.  There are also Fatal (?) errors that scroll past before this hang, and I'd like to deny them as well as the aha1542 and aha152x.  However, I've tried these formats and none allow me to deny multiple modules; always only the last listed module is denied:

nuke denymods="aha1542,aha154x,eata"
nuke denymods="aha1542","aha154x","eata"
nuke denymods=aha1542 denymods=aha154x denymods=eata

What would be the correct way to deny multiple module from loading, without doing this on the originating server?  (I already have the disk image created).

The FAQ only tells how to deny a single module and if there is a solution, this should be updated
http://trac.mondorescue.org/wiki/FAQ#Q36WhyismySLES10restoreblockingatfdomain

Thanks all for the help,

Jeffrey

------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2dcopy1
_______________________________________________
Mondo-devel mailing list
Mon...@li...
https://lists.sourceforge.net/lists/listinfo/mondo-devel

> > > I've tried a few of the Debian 6 packages available but I cannot seem to get this working.
> > > FATAL: Error inserting aha1542

> > Could you send the mondoarchive.log and the mindi.log in gzip format of 
> > that error please?
Not really... the hanging freezes my machine, where would I get that log if I cannot interact?

> Yes, and try to deny that module by booting with denymods="aha1542"
Thanks for the tip; while that 
nuke denymods="aha1542"
works to deny that module, I get hung at the next module, aha152x.  There are also Fatal (?) errors that scroll past before this hang, and I'd like to deny them as well as the aha1542 and aha152x.  However, I've tried these formats and none allow me to deny multiple modules; always only the last listed module is denied:

nuke denymods="aha1542,aha154x,eata"
nuke denymods="aha1542","aha154x","eata"
nuke denymods=aha1542 denymods=aha154x denymods=eata

What would be the correct way to deny multiple module from loading, without doing this on the originating server?  (I already have the disk image created).

The FAQ only tells how to deny a single module and if there is a solution, this should be updated
http://trac.mondorescue.org/wiki/FAQ#Q36WhyismySLES10restoreblockingatfdomain

Thanks all for the help,

Jeffrey

Hello everybody:


        Sorry, if it is not the forum.

I try to implement a solution based in Mondo. I have a machine that boot
from a SAN disk, and I do a mondo of this disk. After, I create a new SAN
disk and I try to recover the mondo in the new disk and boot from it, but
all my test failed.

I would like know if it is possible with mondo and if somebody tried it and
he did it successfully



Thanks in advanced


Luis

Radek <radek_horak <at> cz.ibm.com> writes:


Hi again,

last update - problem is solved.
Cause - McAfee Linux Shield (nails) on-access scan.
After disabling on-access scan in the antivirus, mondo completed fine.
Anyway, thanks for your help, appreciate it.

Hello Alan,

I had the same pb with Dell PE2950 computers and their broadcom NIC that is not recognized 
by mindi.
A solution I found (exposed here http://c-net.fr/info/?doc=Mondo_Rescue,_la_sauvegarde_sous_Linux#restauration
but in french) was to boot with a ubuntu live CD (I had network with it), mount the NFS where data was saved, 
install mondo packages and restore data using mondorestore.

Bruno

Le 26 sept. 2011 à 22:41, Alan Ridgeway a écrit :

> Hello Maillist
>  
> I am new to Mondo Rescue. I am hoping the issue is my error (quicker to solve)   ;)
>  
> When I run mindi, it can create a .iso file. When I boot it in a VM, it comes up and offers a prompt, but the command “interactive” does not work.
> Nor can I see any network info when I run “ifconfig”. If I try to NFS mount to an exported filesystem, it can not find another machine due
> to lack of network access. Hence I do not have a way to get mondorestore.log off the running mindi. I also don’t have a way to restore my backup from NFS.
>  
> When I created mindi, I used the local kernel. The tool can’t find the mindi kernel and when I searched the web site, it was not clear if I need to use the
> source version (which does not match the file extension the tool is looking for. I guess I could rename it.) or the rpm version (which I could not find).
>  
> Relevant OSes:  Red Hat 5.6 and CentOS 5.6 both as VirtualBox 4.1.0 VMs.
> I created mindi .iso file on one VM and try to restore to a second VM with no OS.
>  
> Mindi built using the Red Hat/CentOS kernel (was not sure what mindi kernel to use).
>  
> The following is installed on both:
> afio-2.5-1.rhel5.x86_64.rpm
> buffer-1.19-4.rhel5.x86_64.rpm
> mindi-2.0.7.8-1.rhel5.x86_64.rpm
> mindi-busybox-1.7.3-1.rhel5.x86_64.rpm
> mondo-2.2.9.7-1.rhel5.x86_64.rpm
>  
> This is not installed. Is it important ? If so, what does it do ?
> linuxcoe-sd-base-4.1-1.rhel5.noarch.rpm
>  
> I attached the files mindi-red-hat.log, mindi-centos.log, and mondorescue-centos.log.
> I only ran a backup of centos. That backup when to a NFS server host by Windows 2008 R2.
> I didn’t bother with a backup of Red Hat since I only used it for the moment to see if I would reproduce the error
> with mindi (which I did).
>  
> Any guidance is appreciated.
> Alan
>  
>  
>  
>  
>  
>  
>  
> <mindi-centos.log><mindi-red-hat.log><mondoarchive-centos.log>------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure contains a
> definitive record of customers, application performance, security
> threats, fraudulent activity and more. Splunk takes this data and makes
> sense of it. Business sense. IT sense. Common sense.
> http://p.sf.net/sfu/splunk-d2dcopy1_______________________________________________
> Mondo-devel mailing list
> Mon...@li...
> https://lists.sourceforge.net/lists/listinfo/mondo-devel

Resent in pure text mode so you can actually read the text in the body and not as an attachment.

From: Alan Ridgeway 
Sent: Monday, September 26, 2011 1:42 PM
To: 'mon...@li...'
Subject: Red Hat/CentOS 5.6, mindi boots, but no functional scripts, no network

Hello Maillist

I am new to Mondo Rescue. I am hoping the issue is my error (quicker to solve)   ;)

When I run mindi, it can create a .iso file. When I boot it in a VM, it comes up and offers a prompt, but the command "interactive" does not work.
Nor can I see any network info when I run "ifconfig". If I try to NFS mount to an exported filesystem, it can not find another machine due
to lack of network access. Hence I do not have a way to get mondorestore.log off the running mindi. I also don't have a way to restore my backup from NFS.

When I created mindi, I used the local kernel. The tool can't find the mindi kernel and when I searched the web site, it was not clear if I need to use the
source version (which does not match the file extension the tool is looking for. I guess I could rename it.) or the rpm version (which I could not find).

Relevant OSes:  Red Hat 5.6 and CentOS 5.6 both as VirtualBox 4.1.0 VMs.
I created mindi .iso file on one VM and try to restore to a second VM with no OS.

Mindi built using the Red Hat/CentOS kernel (was not sure what mindi kernel to use).

The following is installed on both:
afio-2.5-1.rhel5.x86_64.rpm
buffer-1.19-4.rhel5.x86_64.rpm
mindi-2.0.7.8-1.rhel5.x86_64.rpm
mindi-busybox-1.7.3-1.rhel5.x86_64.rpm
mondo-2.2.9.7-1.rhel5.x86_64.rpm

This is not installed. Is it important ? If so, what does it do ?
linuxcoe-sd-base-4.1-1.rhel5.noarch.rpm

I attached the files mindi-red-hat.log, mindi-centos.log, and mondorescue-centos.log.
I only ran a backup of centos. That backup when to a NFS server host by Windows 2008 R2.
I didn't bother with a backup of Red Hat since I only used it for the moment to see if I would reproduce the error
with mindi (which I did).

Any guidance is appreciated.
Alan 

Just sent them. Please let me know if/when you get them.

aep

-----Original Message-----
From: Bruno Cornec [mailto:Bru...@hp...] 
Sent: Monday, September 26, 2011 8:28 AM
To: Mondo mailing list
Subject: Re: [Mondo-devel] Mondorescue stops running on a RHEL5.6 system

Alan Pittman said on Mon, Sep 26, 2011 at 11:33:46AM +0000:

>   The mondoarchive log file is gzip'ed. 

Waoh ! Ok. So send it to me privately then, as I don't want some people
of the mailing list paying for their mail to download something which
will be useless for them.

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2dcopy1
_______________________________________________
Mondo-devel mailing list
Mon...@li...
https://lists.sourceforge.net/lists/listinfo/mondo-devel

Alan Pittman said on Mon, Sep 26, 2011 at 11:33:46AM +0000:

>   The mondoarchive log file is gzip'ed. 

Waoh ! Ok. So send it to me privately then, as I don't want some people
of the mailing list paying for their mail to download something which
will be useless for them.

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

Bruno,
  The mondoarchive log file is gzip'ed. 

Alan

-----Original Message-----
From: Bruno Cornec [mailto:Bru...@hp...] 
Sent: Friday, September 23, 2011 5:56 PM
To: Mondo mailing list
Subject: Re: [Mondo-devel] Mondorescue stops running on a RHEL5.6 system

Alan Pittman said on Fri, Sep 23, 2011 at 11:41:27AM +0000:

> I resubmitted this e-mail with the logs attached and got the following reply:
>     Message body is too big: 3400990 bytes with a limit of 400 KB

What about gziping the file ?
(Do not use zip/winzip please, makes it harder for me. I can see
gzip/bzip inline in mutt, not zip)

Bruno.
-- 
Open Source & Linux Profession Lead EMEA           / http://opensource.hp.com
HP/Intel/Red Hat Open Source Solutions Initiative  / http://www.hpintelco.net
http://www.HyPer-Linux.org  http://mondorescue.org http://project-builder.org
La musique ancienne?  http://www.musique-ancienne.org http://www.medieval.org

------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2dcopy2
_______________________________________________
Mondo-devel mailing list
Mon...@li...
https://lists.sourceforge.net/lists/listinfo/mondo-devel

Bruno Cornec <Bruno.Cornec <at> hp.com> writes:

> 
> Radek Horak said on Fri, Sep 16, 2011 at 01:14:35PM +0200:
> 
> > I have a problem with mondo backup on RHEL system.
> > Error reported is :
> > FATAL ERROR. Cannot untar dev-entries.tgz
> 
> Making 32768KB boot disk...............cp: cannot open
> `/usr/lib64/mindi/rootfs/ataraid.tgz' for reading: Permission denied
> cp: cannot open `/usr/lib64/mindi/rootfs/cciss.tgz' for reading:
> Permission denied
> cp: cannot open `/usr/lib64/mindi/rootfs/dev/dev-entries.tgz' for
> reading: Permission denied
> cp: cannot open `/usr/lib64/mindi/rootfs/dm.tgz' for reading: Permission
> denied
> cp: cannot open `/usr/lib64/mindi/rootfs/ida.tgz' for reading:
> Permission denied
> cp: cannot open `/usr/lib64/mindi/rootfs/nst.tgz' for reading:
> Permission denied
> cp: cannot open `/usr/lib64/mindi/rootfs/raw.tgz' for reading:
> Permission denied
> cp: cannot open `/usr/lib64/mindi/rootfs/rd.tgz' for reading: Permission
> denied
> cp: cannot open `/usr/lib64/mindi/rootfs/vc.tgz' for reading: Permission
> denied
> tar: dev-entries.tgz: Cannot open: No such file or directory
> 
> Fix permissions. Or use correct packages. You need to lauch mondoarchive
> as root.
> 
> Also the syntaxt for -E changed so you need to use a | to separate
> directories.
> 
> HTH,
> Bruno.

Hello,
I'm sorry for late reply.
Of course mondoarchive runs under root user, the | separating excludes is not a
problem. 
Fix permissions and correct packages - I already tried this before, but the
files reporting cp and tar problems already have read permissions, same error is
reported even when I set the permissions to 777 for all of the *.tgz files in
/usr/lib64/mindi/rootfs/ and also the file
/usr/lib64/mindi/rootfs/dev/dev-entries.tgz.
Regarding the packages, there is following installed:
mondo-2.2.9.7-1.rhel5
mindi-busybox-1.18.3-3.rhel5
mindi-2.0.7.8-1.rhel5