Re: Questions related to Container pools

[Del Hoobler]

Hi Rick,

We don't have a time line defined for this yet. We have stabilized it for 
now. We encourage customers to look at the more modern and efficient ways 
to solve the use cases that virtual volumes were originally created for.


Del



"ADSM: Dist Stor Manager"  wrote on 05/05/2017 
11:06:43 AM:

> From: Rick Saylor 
> To: ADSM-L@VM.MARIST.EDU
> Date: 05/05/2017 11:07 AM
> Subject: Re: Questions related to Container pools
> Sent by: "ADSM: Dist Stor Manager" 
> 
> Del,
> 
> Can you give us an rough idea of when or, more specifically, which
> release of Spectrum Protect will no longer have virtual volume support?
> 
> Thanks,
> Rick Saylor
> Austin Community College
> 
> On 5/5/2017 5:41 AM, Del Hoobler wrote:
> > Hi Joerg,
> >
> > The original purpose of virtual volumes was to simplify library use 
and
> > sharing in an age where that was difficult. There are significantly 
better
> > ways to solve the problems virtual volumes addressed with capabilities
> > such as fiber channel, shared libraries (library virtualization), and 
even
> > cloud. There are limitations on virtual volumes today and no current 
plans
> > to carry them forward.
> >
> > Del
> >
> > 
> >
> >
> > "ADSM: Dist Stor Manager"  wrote on 05/04/2017
> > 01:38:56 PM:
> >
> >> From: "J. Pohlmann" 
> >> To: ADSM-L@VM.MARIST.EDU
> >> Date: 05/04/2017 01:39 PM
> >> Subject: Re: Questions related to Container pools
> >> Sent by: "ADSM: Dist Stor Manager" 
> >>
> >> Hi Del. W,r,t  Arnaud's q1 - perhaps you could convince your 
colleagues
> > to
> >> also open container storage pools for virtual volume (that is the
> > archive
> >> objects at the target server). I have some folks that like the idea 
of
> >> backing up the database to VVs and also to produce the recovery plan
> > file on
> >> VVs, That way, the DRM set spcifications automate the retention of 
DBB
> > and
> >> RFP objects. Right now, they are using file device class volumes to
> > store
> >> the archive objects that back the VVs.
> >>
> >> Best regards,
> >> Joerg Pohlmann
> >>
> >> -Original Message-
> >> From: ADSM: Dist Stor Manager [mailto:ADSM-L@VM.MARIST.EDU] On Behalf 
Of
> > Del
> >> Hoobler
> >> Sent: Thursday, May 04, 2017 10:26
> >> To: ADSM-L@VM.MARIST.EDU
> >> Subject: Re: [ADSM-L] Questions related to Container pools
> >>
> >> Hi Arnaud,
> >>
> >> For question #1 ... The reason the information is in the document is
> > because
> >> Spectrum Protect restricts it. Stay tuned though, this could be 
allowed
> >> soon.
> >>
> >> For question #2 ... The only reason we recommend using 1 pool is
> > strictly
> >> for the best possible deduplication rates AND the because container
> > pools
> >> can grow much larger in size than its legacy counterparts. For legacy
> > pools,
> >> the housekeeping that was required on the pool made it impossible to
> > grow a
> >> pool too large (identify, reclamations, etc..). With container pools,
> > this
> >> is no longer an issue, so we wanted to call that out. If you want to
> > split
> >> container pools up, knowing the potential dedup rates fallout, that's
> >> perfectly fine ... there are no other repercussions/issues. We have a
> > lot of
> >> customers who have done this, primarily around the partitioning of 
data
> >> sources (VE, TDPs, BA, etc.)
> >>
> >> Thank you,
> >>
> >> Del
> >>
> >> =
> >>
> >>
> >> "ADSM: Dist Stor Manager"  wrote on 05/03/2017
> >> 10:47:52 AM:
> >>
> >>> From: PAC Brion Arnaud 
> >>> To: ADSM-L@VM.MARIST.EDU
> >>> Date: 05/03/2017 10:49 AM
> >>> Subject: Questions related to Container pools Sent by: "ADSM: Dist
> >>> Stor Manager" 
> >>>
> >>> Hi Team !
> >>>
> >>> IBM recently released an interesting document summarizing best
> >>> practices with regards to container storage pools (https://
> >>> 
www.ibm.com/developerworks/community/wikis/home?lang=en#!/wiki/Tivoli
> >>> Storage Manager/page/Container Pool Best Practices ) and some of the
> &

Re: Questions related to Container pools

[Rick Saylor]

Del,

Can you give us an rough idea of when or, more specifically, which
release of Spectrum Protect will no longer have virtual volume support?

Thanks,
Rick Saylor
Austin Community College

On 5/5/2017 5:41 AM, Del Hoobler wrote:

Hi Joerg,

The original purpose of virtual volumes was to simplify library use and
sharing in an age where that was difficult. There are significantly better
ways to solve the problems virtual volumes addressed with capabilities
such as fiber channel, shared libraries (library virtualization), and even
cloud. There are limitations on virtual volumes today and no current plans
to carry them forward.

Del




"ADSM: Dist Stor Manager"  wrote on 05/04/2017
01:38:56 PM:


From: "J. Pohlmann" 
To: ADSM-L@VM.MARIST.EDU
Date: 05/04/2017 01:39 PM
Subject: Re: Questions related to Container pools
Sent by: "ADSM: Dist Stor Manager" 

Hi Del. W,r,t  Arnaud's q1 - perhaps you could convince your colleagues

to

also open container storage pools for virtual volume (that is the

archive

objects at the target server). I have some folks that like the idea of
backing up the database to VVs and also to produce the recovery plan

file on

VVs, That way, the DRM set spcifications automate the retention of DBB

and

RFP objects. Right now, they are using file device class volumes to

store

the archive objects that back the VVs.

Best regards,
Joerg Pohlmann

-Original Message-
From: ADSM: Dist Stor Manager [mailto:ADSM-L@VM.MARIST.EDU] On Behalf Of

Del

Hoobler
Sent: Thursday, May 04, 2017 10:26
To: ADSM-L@VM.MARIST.EDU
Subject: Re: [ADSM-L] Questions related to Container pools

Hi Arnaud,

For question #1 ... The reason the information is in the document is

because

Spectrum Protect restricts it. Stay tuned though, this could be allowed
soon.

For question #2 ... The only reason we recommend using 1 pool is

strictly

for the best possible deduplication rates AND the because container

pools

can grow much larger in size than its legacy counterparts. For legacy

pools,

the housekeeping that was required on the pool made it impossible to

grow a

pool too large (identify, reclamations, etc..). With container pools,

this

is no longer an issue, so we wanted to call that out. If you want to

split

container pools up, knowing the potential dedup rates fallout, that's
perfectly fine ... there are no other repercussions/issues. We have a

lot of

customers who have done this, primarily around the partitioning of data
sources (VE, TDPs, BA, etc.)

Thank you,

Del

=


"ADSM: Dist Stor Manager"  wrote on 05/03/2017
10:47:52 AM:


From: PAC Brion Arnaud 
To: ADSM-L@VM.MARIST.EDU
Date: 05/03/2017 10:49 AM
Subject: Questions related to Container pools Sent by: "ADSM: Dist
Stor Manager" 

Hi Team !

IBM recently released an interesting document summarizing best
practices with regards to container storage pools (https://
www.ibm.com/developerworks/community/wikis/home?lang=en#!/wiki/Tivoli
Storage Manager/page/Container Pool Best Practices ) and some of the
information in it triggered two questions, which I would like some
Spectrum Protect insider (Del ?) or anyone having good knowledge of
this to answer ...

First question :  page 14 of the PDF document, chapter 1.4 , states
that it is not appropriate to use container pools in the case of NDMP
backups. What is the reason for it ? My understanding is that it is
possible to make use of NDMP without a tape based storage pool, thus I
don't get the point here ...

Second question : several references in the book are seeming to stress
that one should make use of ONE SINGLE container storage pool for a
whole TSM server (chapter 1.2.5.1.2 page 13, chapter 1.5.1 page 15). I
do understand that deduplication is made at storage pool level, and
that segregating backup data in several storage pools will weaken
deduplication rates, but are there some other reasons which are not
explained in the book, that would justify the use of only ONE
container pool (like more hammering on the TSM DB during backup times
if we make use of several storage pools, or others I did not think
about) We plan to build a new TSM environment which will be based only
on container storage pool(s ?), and I feel kind of uncomfortable to
send all of my data in the same bucket (less granularity for
reporting, auditing, protecting and so on ...).  Does someone have
arguments (pro or cons) or experience to share about this ?

Thanks in advance for your feedback !

Cheers.

Arnaud





**

Backup and Recovery Systems Administrator Panalpina Management Ltd.,
Basle, Switzerland, CIT Department Viadukstrasse 42, P.O. Box 4002
Basel/CH
Phone: +41 (61) 226 11 11, FAX: +41 (61) 226 17 01
Direct: +41 (61) 226 19 78
e-mail: a

Re: Questions related to Container pools

[Del Hoobler]

Hi Joerg,

The original purpose of virtual volumes was to simplify library use and 
sharing in an age where that was difficult. There are significantly better 
ways to solve the problems virtual volumes addressed with capabilities 
such as fiber channel, shared libraries (library virtualization), and even 
cloud. There are limitations on virtual volumes today and no current plans 
to carry them forward.

Del




"ADSM: Dist Stor Manager"  wrote on 05/04/2017 
01:38:56 PM:

> From: "J. Pohlmann" 
> To: ADSM-L@VM.MARIST.EDU
> Date: 05/04/2017 01:39 PM
> Subject: Re: Questions related to Container pools
> Sent by: "ADSM: Dist Stor Manager" 
> 
> Hi Del. W,r,t  Arnaud's q1 - perhaps you could convince your colleagues 
to
> also open container storage pools for virtual volume (that is the 
archive
> objects at the target server). I have some folks that like the idea of
> backing up the database to VVs and also to produce the recovery plan 
file on
> VVs, That way, the DRM set spcifications automate the retention of DBB 
and
> RFP objects. Right now, they are using file device class volumes to 
store
> the archive objects that back the VVs.
> 
> Best regards,
> Joerg Pohlmann
> 
> -Original Message-
> From: ADSM: Dist Stor Manager [mailto:ADSM-L@VM.MARIST.EDU] On Behalf Of 
Del
> Hoobler
> Sent: Thursday, May 04, 2017 10:26
> To: ADSM-L@VM.MARIST.EDU
> Subject: Re: [ADSM-L] Questions related to Container pools
> 
> Hi Arnaud,
> 
> For question #1 ... The reason the information is in the document is 
because
> Spectrum Protect restricts it. Stay tuned though, this could be allowed
> soon.
> 
> For question #2 ... The only reason we recommend using 1 pool is 
strictly
> for the best possible deduplication rates AND the because container 
pools
> can grow much larger in size than its legacy counterparts. For legacy 
pools,
> the housekeeping that was required on the pool made it impossible to 
grow a
> pool too large (identify, reclamations, etc..). With container pools, 
this
> is no longer an issue, so we wanted to call that out. If you want to 
split
> container pools up, knowing the potential dedup rates fallout, that's
> perfectly fine ... there are no other repercussions/issues. We have a 
lot of
> customers who have done this, primarily around the partitioning of data
> sources (VE, TDPs, BA, etc.)
> 
> Thank you,
> 
> Del
> 
> =
> 
> 
> "ADSM: Dist Stor Manager"  wrote on 05/03/2017
> 10:47:52 AM:
> 
> > From: PAC Brion Arnaud 
> > To: ADSM-L@VM.MARIST.EDU
> > Date: 05/03/2017 10:49 AM
> > Subject: Questions related to Container pools Sent by: "ADSM: Dist
> > Stor Manager" 
> >
> > Hi Team !
> >
> > IBM recently released an interesting document summarizing best
> > practices with regards to container storage pools (https://
> > www.ibm.com/developerworks/community/wikis/home?lang=en#!/wiki/Tivoli
> > Storage Manager/page/Container Pool Best Practices ) and some of the
> > information in it triggered two questions, which I would like some
> > Spectrum Protect insider (Del ?) or anyone having good knowledge of
> > this to answer ...
> >
> > First question :  page 14 of the PDF document, chapter 1.4 , states
> > that it is not appropriate to use container pools in the case of NDMP
> > backups. What is the reason for it ? My understanding is that it is
> > possible to make use of NDMP without a tape based storage pool, thus I
> > don't get the point here ...
> >
> > Second question : several references in the book are seeming to stress
> > that one should make use of ONE SINGLE container storage pool for a
> > whole TSM server (chapter 1.2.5.1.2 page 13, chapter 1.5.1 page 15). I
> > do understand that deduplication is made at storage pool level, and
> > that segregating backup data in several storage pools will weaken
> > deduplication rates, but are there some other reasons which are not
> > explained in the book, that would justify the use of only ONE
> > container pool (like more hammering on the TSM DB during backup times
> > if we make use of several storage pools, or others I did not think
> > about) We plan to build a new TSM environment which will be based only
> > on container storage pool(s ?), and I feel kind of uncomfortable to
> > send all of my data in the same bucket (less granularity for
> > reporting, auditing, protecting and so on ...).  Does someone have
> > arguments (pro or cons) or experience to share about this ?
> >
> > Thanks in advance for your f

Re: Questions related to Container pools

[PAC Brion Arnaud]

Hi Del,

Thanks a lot for clear and appreciated feedback !

Cheers.

Arnaud

-Original Message-
From: ADSM: Dist Stor Manager [mailto:ADSM-L@VM.MARIST.EDU] On Behalf Of Del 
Hoobler
Sent: Thursday, May 04, 2017 7:26 PM
To: ADSM-L@VM.MARIST.EDU
Subject: Re: Questions related to Container pools

Hi Arnaud,

For question #1 ... The reason the information is in the document is 
because Spectrum Protect restricts it. Stay tuned though, this could be 
allowed soon.

For question #2 ... The only reason we recommend using 1 pool is strictly 
for the best possible deduplication rates AND the because container pools 
can grow much larger in size than its legacy counterparts. For legacy 
pools, the housekeeping that was required on the pool made it impossible 
to grow a pool too large (identify, reclamations, etc..). With container 
pools, this is no longer an issue, so we wanted to call that out. If you 
want to split container pools up, knowing the potential dedup rates 
fallout, that's perfectly fine ... there are no other 
repercussions/issues. We have a lot of customers who have done this, 
primarily around the partitioning of data sources (VE, TDPs, BA, etc.)

Thank you,

Del

=


"ADSM: Dist Stor Manager"  wrote on 05/03/2017 
10:47:52 AM:

> From: PAC Brion Arnaud 
> To: ADSM-L@VM.MARIST.EDU
> Date: 05/03/2017 10:49 AM
> Subject: Questions related to Container pools
> Sent by: "ADSM: Dist Stor Manager" 
> 
> Hi Team !
> 
> IBM recently released an interesting document summarizing best 
> practices with regards to container storage pools (https://
> www.ibm.com/developerworks/community/wikis/home?lang=en#!/wiki/Tivoli
> Storage Manager/page/Container Pool Best Practices ) and some of the
> information in it triggered two questions, which I would like some 
> Spectrum Protect insider (Del ?) or anyone having good knowledge of 
> this to answer ...
> 
> First question :  page 14 of the PDF document, chapter 1.4 , states 
> that it is not appropriate to use container pools in the case of 
> NDMP backups. What is the reason for it ? My understanding is that 
> it is possible to make use of NDMP without a tape based storage 
> pool, thus I don't get the point here ...
> 
> Second question : several references in the book are seeming to 
> stress that one should make use of ONE SINGLE container storage pool
> for a whole TSM server (chapter 1.2.5.1.2 page 13, chapter 1.5.1 
> page 15). I do understand that deduplication is made at storage pool
> level, and that segregating backup data in several storage pools 
> will weaken deduplication rates, but are there some other reasons 
> which are not explained in the book, that would justify the use of 
> only ONE container pool (like more hammering on the TSM DB during 
> backup times if we make use of several storage pools, or others I 
> did not think about)
> We plan to build a new TSM environment which will be based only on 
> container storage pool(s ?), and I feel kind of uncomfortable to 
> send all of my data in the same bucket (less granularity for 
> reporting, auditing, protecting and so on ...).  Does someone have 
> arguments (pro or cons) or experience to share about this ?
> 
> Thanks in advance for your feedback !
> 
> Cheers.
> 
> Arnaud
> 
> 
**
> Backup and Recovery Systems Administrator
> Panalpina Management Ltd., Basle, Switzerland,
> CIT Department Viadukstrasse 42, P.O. Box 4002 Basel/CH
> Phone: +41 (61) 226 11 11, FAX: +41 (61) 226 17 01
> Direct: +41 (61) 226 19 78
> e-mail: arnaud.br...@panalpina.com<mailto:arnaud.br...@panalpina.com>
> This electronic message transmission contains information from 
> Panalpina and is confidential or privileged. This information is 
> intended only for the person (s) named above. If you are not the 
> intended recipient, any disclosure, copying, distribution or use or 
> any other action based on the contents of this information is 
> strictly prohibited.
> 
> If you receive this electronic transmission in error, please notify 
> the sender by e-mail, telephone or fax at the numbers listed above. 
Thank you.
> 
**
> 

Re: Questions related to Container pools

[J. Pohlmann]

Hi Del. W,r,t  Arnaud's q1 - perhaps you could convince your colleagues to
also open container storage pools for virtual volume (that is the archive
objects at the target server). I have some folks that like the idea of
backing up the database to VVs and also to produce the recovery plan file on
VVs, That way, the DRM set spcifications automate the retention of DBB and
RFP objects. Right now, they are using file device class volumes to store
the archive objects that back the VVs.

Best regards,
Joerg Pohlmann

-Original Message-
From: ADSM: Dist Stor Manager [mailto:ADSM-L@VM.MARIST.EDU] On Behalf Of Del
Hoobler
Sent: Thursday, May 04, 2017 10:26
To: ADSM-L@VM.MARIST.EDU
Subject: Re: [ADSM-L] Questions related to Container pools

Hi Arnaud,

For question #1 ... The reason the information is in the document is because
Spectrum Protect restricts it. Stay tuned though, this could be allowed
soon.

For question #2 ... The only reason we recommend using 1 pool is strictly
for the best possible deduplication rates AND the because container pools
can grow much larger in size than its legacy counterparts. For legacy pools,
the housekeeping that was required on the pool made it impossible to grow a
pool too large (identify, reclamations, etc..). With container pools, this
is no longer an issue, so we wanted to call that out. If you want to split
container pools up, knowing the potential dedup rates fallout, that's
perfectly fine ... there are no other repercussions/issues. We have a lot of
customers who have done this, primarily around the partitioning of data
sources (VE, TDPs, BA, etc.)

Thank you,

Del

=


"ADSM: Dist Stor Manager"  wrote on 05/03/2017
10:47:52 AM:

> From: PAC Brion Arnaud 
> To: ADSM-L@VM.MARIST.EDU
> Date: 05/03/2017 10:49 AM
> Subject: Questions related to Container pools Sent by: "ADSM: Dist
> Stor Manager" 
>
> Hi Team !
>
> IBM recently released an interesting document summarizing best
> practices with regards to container storage pools (https://
> www.ibm.com/developerworks/community/wikis/home?lang=en#!/wiki/Tivoli
> Storage Manager/page/Container Pool Best Practices ) and some of the
> information in it triggered two questions, which I would like some
> Spectrum Protect insider (Del ?) or anyone having good knowledge of
> this to answer ...
>
> First question :  page 14 of the PDF document, chapter 1.4 , states
> that it is not appropriate to use container pools in the case of NDMP
> backups. What is the reason for it ? My understanding is that it is
> possible to make use of NDMP without a tape based storage pool, thus I
> don't get the point here ...
>
> Second question : several references in the book are seeming to stress
> that one should make use of ONE SINGLE container storage pool for a
> whole TSM server (chapter 1.2.5.1.2 page 13, chapter 1.5.1 page 15). I
> do understand that deduplication is made at storage pool level, and
> that segregating backup data in several storage pools will weaken
> deduplication rates, but are there some other reasons which are not
> explained in the book, that would justify the use of only ONE
> container pool (like more hammering on the TSM DB during backup times
> if we make use of several storage pools, or others I did not think
> about) We plan to build a new TSM environment which will be based only
> on container storage pool(s ?), and I feel kind of uncomfortable to
> send all of my data in the same bucket (less granularity for
> reporting, auditing, protecting and so on ...).  Does someone have
> arguments (pro or cons) or experience to share about this ?
>
> Thanks in advance for your feedback !
>
> Cheers.
>
> Arnaud
>
>

**
> Backup and Recovery Systems Administrator Panalpina Management Ltd.,
> Basle, Switzerland, CIT Department Viadukstrasse 42, P.O. Box 4002
> Basel/CH
> Phone: +41 (61) 226 11 11, FAX: +41 (61) 226 17 01
> Direct: +41 (61) 226 19 78
> e-mail: arnaud.br...@panalpina.com
> This electronic message transmission contains information from
> Panalpina and is confidential or privileged. This information is
> intended only for the person (s) named above. If you are not the
> intended recipient, any disclosure, copying, distribution or use or
> any other action based on the contents of this information is strictly
> prohibited.
>
> If you receive this electronic transmission in error, please notify
> the sender by e-mail, telephone or fax at the numbers listed above.
Thank you.
>

**
>

Re: Questions related to Container pools

[Del Hoobler]

Hi Arnaud,

For question #1 ... The reason the information is in the document is 
because Spectrum Protect restricts it. Stay tuned though, this could be 
allowed soon.

For question #2 ... The only reason we recommend using 1 pool is strictly 
for the best possible deduplication rates AND the because container pools 
can grow much larger in size than its legacy counterparts. For legacy 
pools, the housekeeping that was required on the pool made it impossible 
to grow a pool too large (identify, reclamations, etc..). With container 
pools, this is no longer an issue, so we wanted to call that out. If you 
want to split container pools up, knowing the potential dedup rates 
fallout, that's perfectly fine ... there are no other 
repercussions/issues. We have a lot of customers who have done this, 
primarily around the partitioning of data sources (VE, TDPs, BA, etc.)

Thank you,

Del

=


"ADSM: Dist Stor Manager"  wrote on 05/03/2017 
10:47:52 AM:

> From: PAC Brion Arnaud 
> To: ADSM-L@VM.MARIST.EDU
> Date: 05/03/2017 10:49 AM
> Subject: Questions related to Container pools
> Sent by: "ADSM: Dist Stor Manager" 
> 
> Hi Team !
> 
> IBM recently released an interesting document summarizing best 
> practices with regards to container storage pools (https://
> www.ibm.com/developerworks/community/wikis/home?lang=en#!/wiki/Tivoli
> Storage Manager/page/Container Pool Best Practices ) and some of the
> information in it triggered two questions, which I would like some 
> Spectrum Protect insider (Del ?) or anyone having good knowledge of 
> this to answer ...
> 
> First question :  page 14 of the PDF document, chapter 1.4 , states 
> that it is not appropriate to use container pools in the case of 
> NDMP backups. What is the reason for it ? My understanding is that 
> it is possible to make use of NDMP without a tape based storage 
> pool, thus I don't get the point here ...
> 
> Second question : several references in the book are seeming to 
> stress that one should make use of ONE SINGLE container storage pool
> for a whole TSM server (chapter 1.2.5.1.2 page 13, chapter 1.5.1 
> page 15). I do understand that deduplication is made at storage pool
> level, and that segregating backup data in several storage pools 
> will weaken deduplication rates, but are there some other reasons 
> which are not explained in the book, that would justify the use of 
> only ONE container pool (like more hammering on the TSM DB during 
> backup times if we make use of several storage pools, or others I 
> did not think about)
> We plan to build a new TSM environment which will be based only on 
> container storage pool(s ?), and I feel kind of uncomfortable to 
> send all of my data in the same bucket (less granularity for 
> reporting, auditing, protecting and so on ...).  Does someone have 
> arguments (pro or cons) or experience to share about this ?
> 
> Thanks in advance for your feedback !
> 
> Cheers.
> 
> Arnaud
> 
> 
**
> Backup and Recovery Systems Administrator
> Panalpina Management Ltd., Basle, Switzerland,
> CIT Department Viadukstrasse 42, P.O. Box 4002 Basel/CH
> Phone: +41 (61) 226 11 11, FAX: +41 (61) 226 17 01
> Direct: +41 (61) 226 19 78
> e-mail: arnaud.br...@panalpina.com
> This electronic message transmission contains information from 
> Panalpina and is confidential or privileged. This information is 
> intended only for the person (s) named above. If you are not the 
> intended recipient, any disclosure, copying, distribution or use or 
> any other action based on the contents of this information is 
> strictly prohibited.
> 
> If you receive this electronic transmission in error, please notify 
> the sender by e-mail, telephone or fax at the numbers listed above. 
Thank you.
> 
**
>