With AIX you can build your OVA just running a DD to copy the data in your disk device representation to a file, and you can add your OVF to the mix to create a portable boxed LPAR migration file.

I have created something similar with IBM i , but requires an external storage to map the same disks in a Linux box, so you can DDuplicate your disk devices to files.

I am trying to find a way to do something similar to the AIX DD procedure, even when you have internal disks. 

Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file?

This could help on cloud migrations, building DR-as-a-Service scenarios, creating hot-sites, creating an alternative full-system-backup mechanism based on disk copy, and probably more interesting stuff.

------------------------------
Diego KESSELMAN BARRIONUEVO
ESSELWARE Soluciones, SA de CV
dkesselman@esselware.com
------------------------------

Dear Diego

>>>> Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file? <<<<

In IBM i, we have this thing called "save file" object to which you can save anything in IBM i.  A save file object can be FTPed in binary mode to another IBM i system and its data restored to the target IBM i. It can also be FTPed to any other OS's flat file system but cannot be used in any meaningful way. IBM i also provides a command SAVSTG which you can save the sector-by-sector data (only of permanent objects as opposed to temp objects) in IBM i to a tape device (not sure about to save file).  But I just found that SAVSTG was somehow removed from IBM i as of release 7.2.  I never use this command at all in my 31 years with IBM i. The problem here is that the save file object and IBM i's unique disk sector format are recognized only by IBM i and no other OSs. 

The answer to your question is that someone has to build from scratch a SW utility that understands IBM i unique 520-byte disk sector format used by IBM i internal disk which is possible in principle. But I do not think non-IBM entity will be able to get the needed information from IBM as it is IBM proprietary information.   I see that your approach published in LinkedIn ( IBM i - Migration to the cloud exporting to OVA (no PowerVC required) at https://www.linkedin.com/pulse/ibm-i-migration-cloud-exporting-ova-powervc-diego-kesselman ) and PowerVC are the best one can get now although with the stated requirements.  They are smart as they utilize 512-byte disk format that IBM i was enhanced to accommodate for (in IBM i 6.1) when it supported open storage SAN.  To write a new utility that works on IBM i 520-byte disk format can be a very challenging enterprise.

Or you can try to convince IBM Rochester to totally abandon the use os 520-byte sector format. Since IBM i now can handle 512-bytes format with open storage disk, I imagine it is possible to also change to use 512-byte for internal disk.  If this can happen, things can get more convenient for your purpose.

------------------------------
Education is not the learning of facts but the training of the mind to think. -- Albert Einstein.
------------------------------
Satid S.

Original Message:
Sent: Thu March 09, 2023 09:10 PM
From: Diego KESSELMAN BARRIONUEVO
Subject: IBM i disk device representation on the IFS

With AIX you can build your OVA just running a DD to copy the data in your disk device representation to a file, and you can add your OVF to the mix to create a portable boxed LPAR migration file.

I have created something similar with IBM i , but requires an external storage to map the same disks in a Linux box, so you can DDuplicate your disk devices to files.

I am trying to find a way to do something similar to the AIX DD procedure, even when you have internal disks. 

Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file?

This could help on cloud migrations, building DR-as-a-Service scenarios, creating hot-sites, creating an alternative full-system-backup mechanism based on disk copy, and probably more interesting stuff.

------------------------------
Diego KESSELMAN BARRIONUEVO
ESSELWARE Soluciones, SA de CV
dkesselman@esselware.com
------------------------------

Dear Diego

>>>> Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file? <<<<

In IBM i, we have this thing called "save file" object to which you can save anything in IBM i.  A save file object can be FTPed in binary mode to another IBM i system and its data restored to the target IBM i. It can also be FTPed to any other OS's flat file system but cannot be used in any meaningful way. IBM i also provides a command SAVSTG which you can save the sector-by-sector data (only of permanent objects as opposed to temp objects) in IBM i to a tape device (not sure about to save file).  But I just found that SAVSTG was somehow removed from IBM i as of release 7.2.  I never use this command at all in my 31 years with IBM i. The problem here is that the save file object and IBM i's unique disk sector format are recognized only by IBM i and no other OSs. 

The answer to your question is that someone has to build from scratch a SW utility that understands IBM i unique 520-byte disk sector format used by IBM i internal disk which is possible in principle. But I do not think non-IBM entity will be able to get the needed information from IBM as it is IBM proprietary information.   I see that your approach published in LinkedIn ( IBM i - Migration to the cloud exporting to OVA (no PowerVC required) at https://www.linkedin.com/pulse/ibm-i-migration-cloud-exporting-ova-powervc-diego-kesselman ) and PowerVC are the best one can get now although with the stated requirements.  They are smart as they utilize 512-byte disk format that IBM i was enhanced to accommodate for (in IBM i 6.1) when it supported open storage SAN.  To write a new utility that works on IBM i 520-byte disk format can be a very challenging enterprise.

Or you can try to convince IBM Rochester to totally abandon the use os 520-byte sector format. Since IBM i now can handle 512-bytes format with open storage disk, I imagine it is possible to also change to use 512-byte for internal disk.  If this can happen, things can get more convenient for your purpose.

------------------------------
Education is not the learning of facts but the training of the mind to think. -- Albert Einstein.
------------------------------
Satid S.

Original Message:
Sent: Thu March 09, 2023 09:10 PM
From: Diego KESSELMAN BARRIONUEVO
Subject: IBM i disk device representation on the IFS

With AIX you can build your OVA just running a DD to copy the data in your disk device representation to a file, and you can add your OVF to the mix to create a portable boxed LPAR migration file.

I have created something similar with IBM i , but requires an external storage to map the same disks in a Linux box, so you can DDuplicate your disk devices to files.

I am trying to find a way to do something similar to the AIX DD procedure, even when you have internal disks. 

Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file?

This could help on cloud migrations, building DR-as-a-Service scenarios, creating hot-sites, creating an alternative full-system-backup mechanism based on disk copy, and probably more interesting stuff.

------------------------------
Diego KESSELMAN BARRIONUEVO
ESSELWARE Soluciones, SA de CV
dkesselman@esselware.com
------------------------------

Satid,

IBM i does not support 512-btye block size. It is actually the IBM Storwize SAN storage which was modified to support the IBM i 520-byte block size. 

------------------------------
Steve Pavlichek

Original Message:
Sent: Fri March 10, 2023 02:28 AM
From: Satid Singkorapoom
Subject: IBM i disk device representation on the IFS

Dear Diego

>>>> Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file? <<<<

In IBM i, we have this thing called "save file" object to which you can save anything in IBM i.  A save file object can be FTPed in binary mode to another IBM i system and its data restored to the target IBM i. It can also be FTPed to any other OS's flat file system but cannot be used in any meaningful way. IBM i also provides a command SAVSTG which you can save the sector-by-sector data (only of permanent objects as opposed to temp objects) in IBM i to a tape device (not sure about to save file).  But I just found that SAVSTG was somehow removed from IBM i as of release 7.2.  I never use this command at all in my 31 years with IBM i. The problem here is that the save file object and IBM i's unique disk sector format are recognized only by IBM i and no other OSs. 

The answer to your question is that someone has to build from scratch a SW utility that understands IBM i unique 520-byte disk sector format used by IBM i internal disk which is possible in principle. But I do not think non-IBM entity will be able to get the needed information from IBM as it is IBM proprietary information.   I see that your approach published in LinkedIn ( IBM i - Migration to the cloud exporting to OVA (no PowerVC required) at https://www.linkedin.com/pulse/ibm-i-migration-cloud-exporting-ova-powervc-diego-kesselman ) and PowerVC are the best one can get now although with the stated requirements.  They are smart as they utilize 512-byte disk format that IBM i was enhanced to accommodate for (in IBM i 6.1) when it supported open storage SAN.  To write a new utility that works on IBM i 520-byte disk format can be a very challenging enterprise.

Or you can try to convince IBM Rochester to totally abandon the use os 520-byte sector format. Since IBM i now can handle 512-bytes format with open storage disk, I imagine it is possible to also change to use 512-byte for internal disk.  If this can happen, things can get more convenient for your purpose.

------------------------------
Education is not the learning of facts but the training of the mind to think. -- Albert Einstein.
------------------------------
Satid S.

Original Message:
Sent: Thu March 09, 2023 09:10 PM
From: Diego KESSELMAN BARRIONUEVO
Subject: IBM i disk device representation on the IFS

With AIX you can build your OVA just running a DD to copy the data in your disk device representation to a file, and you can add your OVF to the mix to create a portable boxed LPAR migration file.

I have created something similar with IBM i , but requires an external storage to map the same disks in a Linux box, so you can DDuplicate your disk devices to files.

I am trying to find a way to do something similar to the AIX DD procedure, even when you have internal disks. 

Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file?

This could help on cloud migrations, building DR-as-a-Service scenarios, creating hot-sites, creating an alternative full-system-backup mechanism based on disk copy, and probably more interesting stuff.

------------------------------
Diego KESSELMAN BARRIONUEVO
ESSELWARE Soluciones, SA de CV
dkesselman@esselware.com
------------------------------

Dear Steve

Thanks for your information which is quite a surprise for me. I hope your information is accurate. All the information at the time this support was rolled out was mentioned with regard to IBM i 6.1 which made me think the support was from IBM i itself. 

Here is one sample from the defunct IBM Systems Magazine I still have : 

[QUOTE]

IBM i has been able to attach to 512-byte open storage since early 2008 with the 6.1 release. So far, that connection has required employing VIO server to virtualize open-storage volumes to IBM i, with the IBM i VSCSI client driver providing the "conversion" from 512 bytes to 520 bytes. This conversion involves using one extra 512-byte disk sector for every eight and storing the additional eight bytes from all eight sectors in the ninth.

[UNQUOTE]

Here is another from an IBM redbook on IBM i and StorWize : 

I hope you agree with me that both these samples were written in a way that makes readers think the support is from IBM i itself and therefore a surprise on your revelation! 

Another point is that virtual disk units created from VIOS's internal disk also have 512-byte sector. If your information is accurate, it implies VIOS also has this modified capability as well. 

------------------------------
Education is not the learning of facts but the training of the mind to think. -- Albert Einstein.
------------------------------
Satid S.

Original Message:
Sent: Mon March 13, 2023 08:05 AM
From: Steve Pavlichek
Subject: IBM i disk device representation on the IFS

Satid,

IBM i does not support 512-btye block size. It is actually the IBM Storwize SAN storage which was modified to support the IBM i 520-byte block size. 

------------------------------
Steve Pavlichek

Original Message:
Sent: Fri March 10, 2023 02:28 AM
From: Satid Singkorapoom
Subject: IBM i disk device representation on the IFS

Dear Diego

>>>> Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file? <<<<

In IBM i, we have this thing called "save file" object to which you can save anything in IBM i.  A save file object can be FTPed in binary mode to another IBM i system and its data restored to the target IBM i. It can also be FTPed to any other OS's flat file system but cannot be used in any meaningful way. IBM i also provides a command SAVSTG which you can save the sector-by-sector data (only of permanent objects as opposed to temp objects) in IBM i to a tape device (not sure about to save file).  But I just found that SAVSTG was somehow removed from IBM i as of release 7.2.  I never use this command at all in my 31 years with IBM i. The problem here is that the save file object and IBM i's unique disk sector format are recognized only by IBM i and no other OSs. 

The answer to your question is that someone has to build from scratch a SW utility that understands IBM i unique 520-byte disk sector format used by IBM i internal disk which is possible in principle. But I do not think non-IBM entity will be able to get the needed information from IBM as it is IBM proprietary information.   I see that your approach published in LinkedIn ( IBM i - Migration to the cloud exporting to OVA (no PowerVC required) at https://www.linkedin.com/pulse/ibm-i-migration-cloud-exporting-ova-powervc-diego-kesselman ) and PowerVC are the best one can get now although with the stated requirements.  They are smart as they utilize 512-byte disk format that IBM i was enhanced to accommodate for (in IBM i 6.1) when it supported open storage SAN.  To write a new utility that works on IBM i 520-byte disk format can be a very challenging enterprise.

Or you can try to convince IBM Rochester to totally abandon the use os 520-byte sector format. Since IBM i now can handle 512-bytes format with open storage disk, I imagine it is possible to also change to use 512-byte for internal disk.  If this can happen, things can get more convenient for your purpose.

------------------------------
Education is not the learning of facts but the training of the mind to think. -- Albert Einstein.
------------------------------
Satid S.

Original Message:
Sent: Thu March 09, 2023 09:10 PM
From: Diego KESSELMAN BARRIONUEVO
Subject: IBM i disk device representation on the IFS

With AIX you can build your OVA just running a DD to copy the data in your disk device representation to a file, and you can add your OVF to the mix to create a portable boxed LPAR migration file.

I have created something similar with IBM i , but requires an external storage to map the same disks in a Linux box, so you can DDuplicate your disk devices to files.

I am trying to find a way to do something similar to the AIX DD procedure, even when you have internal disks. 

Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file?

This could help on cloud migrations, building DR-as-a-Service scenarios, creating hot-sites, creating an alternative full-system-backup mechanism based on disk copy, and probably more interesting stuff.

------------------------------
Diego KESSELMAN BARRIONUEVO
ESSELWARE Soluciones, SA de CV
dkesselman@esselware.com
------------------------------

I guess I should not have been multitasking when I replied. You are correct, 

Dear Steve

Thanks for your information which is quite a surprise for me. I hope your information is accurate. All the information at the time this support was rolled out was mentioned with regard to IBM i 6.1 which made me think the support was from IBM i itself. 

Here is one sample from the defunct IBM Systems Magazine I still have : 

[QUOTE]

IBM i has been able to attach to 512-byte open storage since early 2008 with the 6.1 release. So far, that connection has required employing VIO server to virtualize open-storage volumes to IBM i, with the IBM i VSCSI client driver providing the "conversion" from 512 bytes to 520 bytes. This conversion involves using one extra 512-byte disk sector for every eight and storing the additional eight bytes from all eight sectors in the ninth.

[UNQUOTE]

Here is another from an IBM redbook on IBM i and StorWize : 

I hope you agree with me that both these samples were written in a way that makes readers think the support is from IBM i itself and therefore a surprise on your revelation! 

Another point is that virtual disk units created from VIOS's internal disk also have 512-byte sector. If your information is accurate, it implies VIOS also has this modified capability as well. 

------------------------------
Education is not the learning of facts but the training of the mind to think. -- Albert Einstein.
------------------------------
Satid S.

Original Message:
Sent: Mon March 13, 2023 08:05 AM
From: Steve Pavlichek
Subject: IBM i disk device representation on the IFS

Satid,

IBM i does not support 512-btye block size. It is actually the IBM Storwize SAN storage which was modified to support the IBM i 520-byte block size. 

------------------------------
Steve Pavlichek

Original Message:
Sent: Fri March 10, 2023 02:28 AM
From: Satid Singkorapoom
Subject: IBM i disk device representation on the IFS

Dear Diego

>>>> Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file? <<<<

In IBM i, we have this thing called "save file" object to which you can save anything in IBM i.  A save file object can be FTPed in binary mode to another IBM i system and its data restored to the target IBM i. It can also be FTPed to any other OS's flat file system but cannot be used in any meaningful way. IBM i also provides a command SAVSTG which you can save the sector-by-sector data (only of permanent objects as opposed to temp objects) in IBM i to a tape device (not sure about to save file).  But I just found that SAVSTG was somehow removed from IBM i as of release 7.2.  I never use this command at all in my 31 years with IBM i. The problem here is that the save file object and IBM i's unique disk sector format are recognized only by IBM i and no other OSs. 

The answer to your question is that someone has to build from scratch a SW utility that understands IBM i unique 520-byte disk sector format used by IBM i internal disk which is possible in principle. But I do not think non-IBM entity will be able to get the needed information from IBM as it is IBM proprietary information.   I see that your approach published in LinkedIn ( IBM i - Migration to the cloud exporting to OVA (no PowerVC required) at https://www.linkedin.com/pulse/ibm-i-migration-cloud-exporting-ova-powervc-diego-kesselman ) and PowerVC are the best one can get now although with the stated requirements.  They are smart as they utilize 512-byte disk format that IBM i was enhanced to accommodate for (in IBM i 6.1) when it supported open storage SAN.  To write a new utility that works on IBM i 520-byte disk format can be a very challenging enterprise.

Or you can try to convince IBM Rochester to totally abandon the use os 520-byte sector format. Since IBM i now can handle 512-bytes format with open storage disk, I imagine it is possible to also change to use 512-byte for internal disk.  If this can happen, things can get more convenient for your purpose.

------------------------------
Education is not the learning of facts but the training of the mind to think. -- Albert Einstein.
------------------------------
Satid S.

Original Message:
Sent: Thu March 09, 2023 09:10 PM
From: Diego KESSELMAN BARRIONUEVO
Subject: IBM i disk device representation on the IFS

With AIX you can build your OVA just running a DD to copy the data in your disk device representation to a file, and you can add your OVF to the mix to create a portable boxed LPAR migration file.

I have created something similar with IBM i , but requires an external storage to map the same disks in a Linux box, so you can DDuplicate your disk devices to files.

I am trying to find a way to do something similar to the AIX DD procedure, even when you have internal disks. 

Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file?

This could help on cloud migrations, building DR-as-a-Service scenarios, creating hot-sites, creating an alternative full-system-backup mechanism based on disk copy, and probably more interesting stuff.

------------------------------
Diego KESSELMAN BARRIONUEVO
ESSELWARE Soluciones, SA de CV
dkesselman@esselware.com
------------------------------

Dear Diego

>>>> Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file? <<<<

In IBM i, we have this thing called "save file" object to which you can save anything in IBM i.  A save file object can be FTPed in binary mode to another IBM i system and its data restored to the target IBM i. It can also be FTPed to any other OS's flat file system but cannot be used in any meaningful way. IBM i also provides a command SAVSTG which you can save the sector-by-sector data (only of permanent objects as opposed to temp objects) in IBM i to a tape device (not sure about to save file).  But I just found that SAVSTG was somehow removed from IBM i as of release 7.2.  I never use this command at all in my 31 years with IBM i. The problem here is that the save file object and IBM i's unique disk sector format are recognized only by IBM i and no other OSs. 

The answer to your question is that someone has to build from scratch a SW utility that understands IBM i unique 520-byte disk sector format used by IBM i internal disk which is possible in principle. But I do not think non-IBM entity will be able to get the needed information from IBM as it is IBM proprietary information.   I see that your approach published in LinkedIn ( IBM i - Migration to the cloud exporting to OVA (no PowerVC required) at https://www.linkedin.com/pulse/ibm-i-migration-cloud-exporting-ova-powervc-diego-kesselman ) and PowerVC are the best one can get now although with the stated requirements.  They are smart as they utilize 512-byte disk format that IBM i was enhanced to accommodate for (in IBM i 6.1) when it supported open storage SAN.  To write a new utility that works on IBM i 520-byte disk format can be a very challenging enterprise.

Or you can try to convince IBM Rochester to totally abandon the use os 520-byte sector format. Since IBM i now can handle 512-bytes format with open storage disk, I imagine it is possible to also change to use 512-byte for internal disk.  If this can happen, things can get more convenient for your purpose.

------------------------------
Education is not the learning of facts but the training of the mind to think. -- Albert Einstein.
------------------------------
Satid S.

Original Message:
Sent: Thu March 09, 2023 09:10 PM
From: Diego KESSELMAN BARRIONUEVO
Subject: IBM i disk device representation on the IFS

With AIX you can build your OVA just running a DD to copy the data in your disk device representation to a file, and you can add your OVF to the mix to create a portable boxed LPAR migration file.

I have created something similar with IBM i , but requires an external storage to map the same disks in a Linux box, so you can DDuplicate your disk devices to files.

I am trying to find a way to do something similar to the AIX DD procedure, even when you have internal disks. 

Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file?

This could help on cloud migrations, building DR-as-a-Service scenarios, creating hot-sites, creating an alternative full-system-backup mechanism based on disk copy, and probably more interesting stuff.

------------------------------
Diego KESSELMAN BARRIONUEVO
ESSELWARE Soluciones, SA de CV
dkesselman@esselware.com
------------------------------

I could have sworn SAVSTG died a long time ago:

https://www.ibm.com/docs/en/i/7.2?topic=ios-save-storage-savstg-command-restore-storage-process-are-no-longer-provided

and on my 7.5 system I get:   Command SAVSTG in library *LIBL not found.

Lots of 7.5 documentation needs updated though.

Dear Diego

>>>> Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file? <<<<

In IBM i, we have this thing called "save file" object to which you can save anything in IBM i.  A save file object can be FTPed in binary mode to another IBM i system and its data restored to the target IBM i. It can also be FTPed to any other OS's flat file system but cannot be used in any meaningful way. IBM i also provides a command SAVSTG which you can save the sector-by-sector data (only of permanent objects as opposed to temp objects) in IBM i to a tape device (not sure about to save file).  But I just found that SAVSTG was somehow removed from IBM i as of release 7.2.  I never use this command at all in my 31 years with IBM i. The problem here is that the save file object and IBM i's unique disk sector format are recognized only by IBM i and no other OSs. 

The answer to your question is that someone has to build from scratch a SW utility that understands IBM i unique 520-byte disk sector format used by IBM i internal disk which is possible in principle. But I do not think non-IBM entity will be able to get the needed information from IBM as it is IBM proprietary information.   I see that your approach published in LinkedIn ( IBM i - Migration to the cloud exporting to OVA (no PowerVC required) at https://www.linkedin.com/pulse/ibm-i-migration-cloud-exporting-ova-powervc-diego-kesselman ) and PowerVC are the best one can get now although with the stated requirements.  They are smart as they utilize 512-byte disk format that IBM i was enhanced to accommodate for (in IBM i 6.1) when it supported open storage SAN.  To write a new utility that works on IBM i 520-byte disk format can be a very challenging enterprise.

Or you can try to convince IBM Rochester to totally abandon the use os 520-byte sector format. Since IBM i now can handle 512-bytes format with open storage disk, I imagine it is possible to also change to use 512-byte for internal disk.  If this can happen, things can get more convenient for your purpose.

------------------------------
Education is not the learning of facts but the training of the mind to think. -- Albert Einstein.
------------------------------
Satid S.

Original Message:
Sent: Thu March 09, 2023 09:10 PM
From: Diego KESSELMAN BARRIONUEVO
Subject: IBM i disk device representation on the IFS

With AIX you can build your OVA just running a DD to copy the data in your disk device representation to a file, and you can add your OVF to the mix to create a portable boxed LPAR migration file.

I have created something similar with IBM i , but requires an external storage to map the same disks in a Linux box, so you can DDuplicate your disk devices to files.

I am trying to find a way to do something similar to the AIX DD procedure, even when you have internal disks. 

Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file?

This could help on cloud migrations, building DR-as-a-Service scenarios, creating hot-sites, creating an alternative full-system-backup mechanism based on disk copy, and probably more interesting stuff.

------------------------------
Diego KESSELMAN BARRIONUEVO
ESSELWARE Soluciones, SA de CV
dkesselman@esselware.com
------------------------------

With AIX you can build your OVA just running a DD to copy the data in your disk device representation to a file, and you can add your OVF to the mix to create a portable boxed LPAR migration file.

I have created something similar with IBM i , but requires an external storage to map the same disks in a Linux box, so you can DDuplicate your disk devices to files.

I am trying to find a way to do something similar to the AIX DD procedure, even when you have internal disks. 

Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file?

This could help on cloud migrations, building DR-as-a-Service scenarios, creating hot-sites, creating an alternative full-system-backup mechanism based on disk copy, and probably more interesting stuff.

With AIX you can build your OVA just running a DD to copy the data in your disk device representation to a file, and you can add your OVF to the mix to create a portable boxed LPAR migration file.

I have created something similar with IBM i , but requires an external storage to map the same disks in a Linux box, so you can DDuplicate your disk devices to files.

I am trying to find a way to do something similar to the AIX DD procedure, even when you have internal disks. 

Does anyone know a way to get a device representation from your disks to DD to a file, or any other mechanism to read disk data and send it to an image file?

This could help on cloud migrations, building DR-as-a-Service scenarios, creating hot-sites, creating an alternative full-system-backup mechanism based on disk copy, and probably more interesting stuff.