From: Frank S. <fs@WPI.EDU> - 2008-01-17 13:49:10
|
Item: Allow configuration of multiple media type capabilities per storage device Date: 17 January 2008 Origin: Frank Sweetser <fs...@wp...> What: Extend the media type configuration for storage devices from a single type to multiple types, including specifying that some are read-only. Why: As various storage types evolve, a single media type per storage device is no longer adequate. For example, LTO-4 drives can read LTO-4, LTO-3, and LTO-2 tapes, but can only write to LTO-4 and LTO-3. Lack of this information in a heterogeneous configuration could lead to failed backups if Bacula selects a non-writable tape from a library for a backup job. Also, the only currently available workaround (lie to Bacula about the media type of some volumes) means that it's often not possible to generate an accurate volume inventory from the catalog. Notes: One simple way to implement this might be redefine the current 'Media Type' directive into a comma separated list, and add another 'Read Only Media Type' directive which lists the media types that device is capable of reading but not writing. This should provide a backwards compatible way of extending the media type configuration. -- Frank Sweetser fs at wpi.edu | For every problem, there is a solution that WPI Senior Network Engineer | is simple, elegant, and wrong. - HL Mencken GPG fingerprint = 6174 1257 129E 0D21 D8D4 E8A3 8E39 29E3 E2E8 8CEC |
From: Kern S. <ke...@si...> - 2008-01-17 14:06:46
|
Hello, I believe that in version 2.2.x part of what you want may be implemented. I have never tested it, but the 2.2.x SD should accept different Media Types within an Autochanger. It will, however, only permit one media type per drive and does not have the concept of read-only. Regards, Kern On Thursday 17 January 2008 14:49, Frank Sweetser wrote: > Item: Allow configuration of multiple media type capabilities per storage > device > > Date: 17 January 2008 > > Origin: Frank Sweetser <fs...@wp...> > > What: Extend the media type configuration for storage devices from a single > type to multiple types, including specifying that some are read-only. > > Why: As various storage types evolve, a single media type per storage > device is no longer adequate. For example, LTO-4 drives can read LTO-4, > LTO-3, and LTO-2 tapes, but can only write to LTO-4 and LTO-3. Lack of > this information in a heterogeneous configuration could lead to failed > backups if Bacula selects a non-writable tape from a library for a backup > job. Also, the only currently available workaround (lie to Bacula about > the media type of some volumes) means that it's often not possible to > generate an accurate volume inventory from the catalog. > > Notes: One simple way to implement this might be redefine the current > 'Media Type' directive into a comma separated list, and add another 'Read > Only Media Type' directive which lists the media types that device is > capable of reading but not writing. This should provide a backwards > compatible way of extending the media type configuration. |
From: Frank S. <fs@WPI.EDU> - 2008-01-17 14:08:32
|
Kern Sibbald wrote: > Hello, > > I believe that in version 2.2.x part of what you want may be implemented. I > have never tested it, but the 2.2.x SD should accept different Media Types > within an Autochanger. It will, however, only permit one media type per So you could support, for example, an autochanger with one LTO-3 drive and one LTO-4 drive? > drive and does not have the concept of read-only. Unfortunately the multiple media types per drive is the important bit =) -- Frank Sweetser fs at wpi.edu | For every problem, there is a solution that WPI Senior Network Engineer | is simple, elegant, and wrong. - HL Mencken GPG fingerprint = 6174 1257 129E 0D21 D8D4 E8A3 8E39 29E3 E2E8 8CEC |
From: Kern S. <ke...@si...> - 2008-01-17 15:40:07
|
On Thursday 17 January 2008 15:08, Frank Sweetser wrote: > Kern Sibbald wrote: > > Hello, > > > > I believe that in version 2.2.x part of what you want may be implemented. > > I have never tested it, but the 2.2.x SD should accept different Media > > Types within an Autochanger. It will, however, only permit one media > > type per > > So you could support, for example, an autochanger with one LTO-3 drive and > one LTO-4 drive? That was what I was hoping to achieve, but as I said, I never tested it. > > > drive and does not have the concept of read-only. > > Unfortunately the multiple media types per drive is the important bit =) Yes, I understand ... |
From: Frank S. <fs@WPI.EDU> - 2008-01-17 16:23:40
|
Kern Sibbald wrote: > On Thursday 17 January 2008 15:08, Frank Sweetser wrote: >> Kern Sibbald wrote: >>> Hello, >>> >>> I believe that in version 2.2.x part of what you want may be implemented. >>> I have never tested it, but the 2.2.x SD should accept different Media >>> Types within an Autochanger. It will, however, only permit one media >>> type per >> So you could support, for example, an autochanger with one LTO-3 drive and >> one LTO-4 drive? > > That was what I was hoping to achieve, but as I said, I never tested it. If I get a chance in the next couple of weeks, I'll see if I can throw together a regression test for it. -- Frank Sweetser fs at wpi.edu | For every problem, there is a solution that WPI Senior Network Engineer | is simple, elegant, and wrong. - HL Mencken GPG fingerprint = 6174 1257 129E 0D21 D8D4 E8A3 8E39 29E3 E2E8 8CEC |
From: Kern S. <ke...@si...> - 2008-01-17 17:13:57
|
On Thursday 17 January 2008 16:41, Frank Sweetser wrote: > Kern Sibbald wrote: > > On Thursday 17 January 2008 15:08, Frank Sweetser wrote: > >> Kern Sibbald wrote: > >>> Hello, > >>> > >>> I believe that in version 2.2.x part of what you want may be > >>> implemented. I have never tested it, but the 2.2.x SD should accept > >>> different Media Types within an Autochanger. It will, however, only > >>> permit one media type per > >> > >> So you could support, for example, an autochanger with one LTO-3 drive > >> and one LTO-4 drive? > > > > That was what I was hoping to achieve, but as I said, I never tested it. > > If I get a chance in the next couple of weeks, I'll see if I can throw > together a regression test for it. Oh, that would be *very* nice. :-) |