Re: [Bacula-users] possible bug in bacula-sd, relating to Type=Files and autochanger

2010-11-15 Thread Philip Brown
On Sat, Nov 13, 2010 at 6:01 PM, Dan Langille d...@langille.org wrote:
 .

 The project needs more people.  Can you help?  It is things like this for
 which patches are required.  :)


I'd be willing to help... the problem is, I dont feel that I
sufficiently understand the subject to write an update.
Sometimes, documenting things that work, when they work for bad
reasons, is worse than no documentation.
Better they would be written by, for example, the person that put that
working snippet into the regression testing tree?

--
Centralized Desktop Delivery: Dell and VMware Reference Architecture
Simplifying enterprise desktop deployment and management using
Dell EqualLogic storage and VMware View: A highly scalable, end-to-end
client virtualization framework. Read more!
http://p.sf.net/sfu/dell-eql-dev2dev
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] possible bug in bacula-sd, relating to Type=Files and autochanger

2010-11-15 Thread Dan Langille
On 11/15/2010 12:38 PM, Philip Brown wrote:
 On Sat, Nov 13, 2010 at 6:01 PM, Dan Langilled...@langille.org  wrote:
 .

 The project needs more people.  Can you help?  It is things like this for
 which patches are required.  :)


 I'd be willing to help... the problem is, I dont feel that I
 sufficiently understand the subject to write an update.
 Sometimes, documenting things that work, when they work for bad
 reasons, is worse than no documentation.

Thing is, the doc project needs someone to take charge and then 
coordinate items.  e.g. establish what needs to be done, then hand out 
the tasks.

 Better they would be written by, for example, the person that put that
 working snippet into the regression testing tree?

In short, definitely not.  :)

In general people are bad at knowing what they didn't know.  They only 
know what they know now. So, it's better for someone who doesn't know 
how something works to write about it. Any gaps in knowledge can be 
obtained by asking.  Or by trying.

And again, you don't have to be doing the writing.  People will take up 
the tasks. They just need someone to tell them what needs doing.


-- 
Dan Langille - http://langille.org/

--
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2  L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today
http://p.sf.net/sfu/msIE9-sfdev2dev
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] possible bug in bacula-sd, relating to Type=Files and autochanger

2010-11-13 Thread Dan Langille
On 11/12/2010 3:21 PM, Philip Brown wrote:
 Martin Simmons writes:

 Have you looked at how the regression tests do it?  In particular:
 http://bacula.git.sourceforge.net/git/gitweb.cgi?p=bacula/bacula;a=blob;f=regress/scripts/bacula-sd-2disk.conf.in;h=7f8f33e9fb1b896aa65db3739dda1d2d367b4154;hb=HEAD

 I don't know if that works, but it is subtly different from your setup.


 WOW!

 The good news is... it seems to work. Thank you so much!

 The bad news is this means that the bacula documentation, and
 shipped examples, are horribly horribly out of date, and just plain
 WRONG!

While I'll disagree with the chosen adjective

The project needs more people.  Can you help?  It is things like this 
for which patches are required.  :)




 I will start by mentioning that I'm using bacula 5.0.3

 the examples/bacula-sd.conf which is autogenerated, has one and only
 one section for a
 filesystem based device, which looks like the following:


 Device {
Name = FileStorage
Media Type = File
Archive Device = /tmp
LabelMedia = yes;   # lets Bacula label unlabeled media
Random Access = Yes;
AutomaticMount = yes;   # when device opened, read it
RemovableMedia = no;
AlwaysOpen = no;
 }


 Interestingly, that works when by itself. But not referenced by an 
 autochanger
 The Device declaration that actually works when used with a virtual
 autochanger, is

 Device {
Name = FileStorage0
Device Type = File
Media Type = Disk
Archive Device = /dir/name
AutomaticMount = yes;
Autochanger = yes
Drive Index = 0
AlwaysOpen = yes;
RemovableMedia = yes;
 }

 Notice that there is a Device Type that seems to be required.. that
 was totally missing.
 Notice also, that Media Type is now specified as Disk, instead of File 
 !!!
 (Hmm. I guess this is free-form and does not strictly matter. But
 its still really confusing and bug-inducing, to have that
 inconsistency!)

 It's really tough to find any reference under
 http://www.bacula.org/5.0.x-manuals/en/main/
 of This is the proper way to set up a Filesystem based storage demon device

 There is no official example of a Disk based device in
 (src)/examples/devices either.


 PS: I think my original message was right, in that there is some kind
 of subtle bug in the storage demon code.
 It seems like, *if Device Type is not specified*,  it defaults to type
 = File, when archive device is a directory, and it is a standalone
 device. But it defaults to type=Tape, when device is associated with
 an autochanger and not otherwise specified.

 Possibly the best way to fix this would be to insist that Device
 Type be always specified explicitly.

 --
 Centralized Desktop Delivery: Dell and VMware Reference Architecture
 Simplifying enterprise desktop deployment and management using
 Dell EqualLogic storage and VMware View: A highly scalable, end-to-end
 client virtualization framework. Read more!
 http://p.sf.net/sfu/dell-eql-dev2dev
 ___
 Bacula-users mailing list
 Bacula-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/bacula-users



-- 
Dan Langille - http://langille.org/

--
Centralized Desktop Delivery: Dell and VMware Reference Architecture
Simplifying enterprise desktop deployment and management using
Dell EqualLogic storage and VMware View: A highly scalable, end-to-end
client virtualization framework. Read more!
http://p.sf.net/sfu/dell-eql-dev2dev
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] possible bug in bacula-sd, relating to Type=Files and autochanger

2010-11-12 Thread Martin Simmons
 On Thu, 11 Nov 2010 16:04:35 -0800, Philip Brown said:
 
 Hello,
 
 I've been wrestling with configuration of a filesystem based
 autochanger, aka the 'disk-changer' script.
 
 I've made some progress, but have hit a point where I believe a bug in
 the core sd libraries (specifically around src/stored/dev.c:549) is
 stopping me from going further.
 
 First of all, I have a regular, non-autochanger device working just
 great. I can successfully label volumes and use them.
 
 I then got the basic autochanger code working, by reading the script,
 and figuring out that most of the examples to use it are now wrong.
 This is because for media type=files, the Device clause expects a
 reference to a directory, for the Archive Device.

I think that last assumption is wrong for a disk-based autochanger.

Have you looked at how the regression tests do it?  In particular:

http://bacula.git.sourceforge.net/git/gitweb.cgi?p=bacula/bacula;a=blob;f=regress/scripts/bacula-sd-2disk.conf.in;h=7f8f33e9fb1b896aa65db3739dda1d2d367b4154;hb=HEAD

I don't know if that works, but it is subtly different from your setup.

__Martin

--
Centralized Desktop Delivery: Dell and VMware Reference Architecture
Simplifying enterprise desktop deployment and management using
Dell EqualLogic storage and VMware View: A highly scalable, end-to-end
client virtualization framework. Read more!
http://p.sf.net/sfu/dell-eql-dev2dev
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] possible bug in bacula-sd, relating to Type=Files and autochanger

2010-11-12 Thread Philip Brown
Martin Simmons writes:

Have you looked at how the regression tests do it?  In particular:
http://bacula.git.sourceforge.net/git/gitweb.cgi?p=bacula/bacula;a=blob;f=regress/scripts/bacula-sd-2disk.conf.in;h=7f8f33e9fb1b896aa65db3739dda1d2d367b4154;hb=HEAD

I don't know if that works, but it is subtly different from your setup.


WOW!

The good news is... it seems to work. Thank you so much!

The bad news is this means that the bacula documentation, and
shipped examples, are horribly horribly out of date, and just plain
WRONG!

I will start by mentioning that I'm using bacula 5.0.3

the examples/bacula-sd.conf which is autogenerated, has one and only
one section for a
filesystem based device, which looks like the following:


Device {
  Name = FileStorage
  Media Type = File
  Archive Device = /tmp
  LabelMedia = yes;   # lets Bacula label unlabeled media
  Random Access = Yes;
  AutomaticMount = yes;   # when device opened, read it
  RemovableMedia = no;
  AlwaysOpen = no;
}


Interestingly, that works when by itself. But not referenced by an autochanger
The Device declaration that actually works when used with a virtual
autochanger, is

Device {
  Name = FileStorage0
  Device Type = File
  Media Type = Disk
  Archive Device = /dir/name
  AutomaticMount = yes;
  Autochanger = yes
  Drive Index = 0
  AlwaysOpen = yes;
  RemovableMedia = yes;
}

Notice that there is a Device Type that seems to be required.. that
was totally missing.
Notice also, that Media Type is now specified as Disk, instead of File !!!
(Hmm. I guess this is free-form and does not strictly matter. But
its still really confusing and bug-inducing, to have that
inconsistency!)

It's really tough to find any reference under
http://www.bacula.org/5.0.x-manuals/en/main/
of This is the proper way to set up a Filesystem based storage demon device

There is no official example of a Disk based device in
(src)/examples/devices either.


PS: I think my original message was right, in that there is some kind
of subtle bug in the storage demon code.
It seems like, *if Device Type is not specified*,  it defaults to type
= File, when archive device is a directory, and it is a standalone
device. But it defaults to type=Tape, when device is associated with
an autochanger and not otherwise specified.

Possibly the best way to fix this would be to insist that Device
Type be always specified explicitly.

--
Centralized Desktop Delivery: Dell and VMware Reference Architecture
Simplifying enterprise desktop deployment and management using
Dell EqualLogic storage and VMware View: A highly scalable, end-to-end
client virtualization framework. Read more!
http://p.sf.net/sfu/dell-eql-dev2dev
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users