|
|
|
date: Mon, 3 Mar 2008 06:39:31 -0500,
group: microsoft.public.exchange.design
back
Re: SCR of CCR'd stores
- This is certainly possible and a common/recommended topology to get high
availability locally (CCR) to protect against server and storage/database
failure, and get site resilience by replicating to a remote SCR target.
- Setting up Storage Groups on a CMS in a clustered environment is no
different than setting it up for a standalone/non-clustered server.
- The only thing you need to consider (and recommend testing in a lab to be
comfortable, before deploying in production), imo, is the path to recovery.
(In addition to planning paths for Storage Group + Database on both source
and target servers.. they need to be identical).
- Understand that SCR target requires activation - there are no automated
failures as in the case of CCR/SCC clusters. SCR is not a cluster.
- If the SCR source and SCR target are clustered (in this case SCR target
should not have a CMS created, just cluster setup and Exchange server
setup), you can use /recoverCMS
- The SCR target can be a single node cluster (consideration: licensing cost
increase).
--
Bharat Suneja
MVP - Exchange
www.zenprise.com
NEW blog location:
exchangepedia.com/blog
----------------------------
"Thinkpad21" wrote in message
news:uQNk%23NSfIHA.4140@TK2MSFTNGP04.phx.gbl...
> Hi,
>
> For site resiliency design, we are going to SCR to another location the
> CCR'd stores on the corporate site.
>
> I have been looking for exact documentation on how to implement this (
> step by step kind of document, not high level design )...
>
> Wondering what would be the easiest way to configure the Mailbox server in
> the resiliency site for SCR ( Standalone or single node cluster ). This
> site has about 80 users that will have a local store on this server as
> well as having the SCR copy of the corporate site that has about 700
> users...
>
> Since the resiliency site will be very small, we were also planning to
> have the HUB/CAS role and Public folders also on this SCR server...
>
> From what I have read, this is all possible, but wondering where I can
> find the implementation steps and also, how to use the SCR copy in case of
> site disaster at corporate site, I need to document the steps required to
> bring the SCR copy live and have users pointing to it...
>
> Thanks !!!
>
> --
>
>
date: Mon, 3 Mar 2008 11:40:44 -0800
author: Bharat Suneja [MVP]
Re: SCR of CCR'd stores
Thanks, I guess my question is, will it be easier to activate SCR if the scr
target was a single node cluster or stand-alone server...
I know that we can use DB portability for the stand-alone scenario, but that
means that Outlook profiles need to be modified if Outlook 2007 is not used.
So if that's the case, that would require quite an effort...
If we go with a SCR that is a single node cluster, than a RecoverCMS would
be needed, but in this case, can the server be also a Hub/CAS and have a
public folder store ?
I have a complete lab infrastructure and test this on it, but was just
wondering.
Thanks !!!
--
--
"Bharat Suneja [MVP]" wrote in message
news:uUXJ0ZWfIHA.4684@TK2MSFTNGP06.phx.gbl...
>- This is certainly possible and a common/recommended topology to get high
>availability locally (CCR) to protect against server and storage/database
>failure, and get site resilience by replicating to a remote SCR target.
> - Setting up Storage Groups on a CMS in a clustered environment is no
> different than setting it up for a standalone/non-clustered server.
> - The only thing you need to consider (and recommend testing in a lab to
> be comfortable, before deploying in production), imo, is the path to
> recovery.
> (In addition to planning paths for Storage Group + Database on both source
> and target servers.. they need to be identical).
> - Understand that SCR target requires activation - there are no automated
> failures as in the case of CCR/SCC clusters. SCR is not a cluster.
> - If the SCR source and SCR target are clustered (in this case SCR target
> should not have a CMS created, just cluster setup and Exchange server
> setup), you can use /recoverCMS
> - The SCR target can be a single node cluster (consideration: licensing
> cost increase).
>
> --
> Bharat Suneja
> MVP - Exchange
> www.zenprise.com
> NEW blog location:
> exchangepedia.com/blog
> ----------------------------
>
>
> "Thinkpad21" wrote in message
> news:uQNk%23NSfIHA.4140@TK2MSFTNGP04.phx.gbl...
>> Hi,
>>
>> For site resiliency design, we are going to SCR to another location the
>> CCR'd stores on the corporate site.
>>
>> I have been looking for exact documentation on how to implement this (
>> step by step kind of document, not high level design )...
>>
>> Wondering what would be the easiest way to configure the Mailbox server
>> in the resiliency site for SCR ( Standalone or single node cluster ).
>> This site has about 80 users that will have a local store on this server
>> as well as having the SCR copy of the corporate site that has about 700
>> users...
>>
>> Since the resiliency site will be very small, we were also planning to
>> have the HUB/CAS role and Public folders also on this SCR server...
>>
>> From what I have read, this is all possible, but wondering where I can
>> find the implementation steps and also, how to use the SCR copy in case
>> of site disaster at corporate site, I need to document the steps required
>> to bring the SCR copy live and have users pointing to it...
>>
>> Thanks !!!
>>
>> --
>>
>>
>
date: Mon, 3 Mar 2008 15:56:15 -0500
author: Thinkpad21
Re: SCR of CCR'd stores
- /RecoverCMS would be the quickest option, imo.
- Unfortunately, this means you can't have Hub/CAS on the same box.
--
Bharat Suneja
MVP - Exchange
www.zenprise.com
NEW blog location:
exchangepedia.com/blog
----------------------------
"Thinkpad21" wrote in message
news:uiUFJFXfIHA.5280@TK2MSFTNGP02.phx.gbl...
> Thanks, I guess my question is, will it be easier to activate SCR if the
> scr target was a single node cluster or stand-alone server...
>
> I know that we can use DB portability for the stand-alone scenario, but
> that means that Outlook profiles need to be modified if Outlook 2007 is
> not used. So if that's the case, that would require quite an effort...
>
> If we go with a SCR that is a single node cluster, than a RecoverCMS would
> be needed, but in this case, can the server be also a Hub/CAS and have a
> public folder store ?
>
> I have a complete lab infrastructure and test this on it, but was just
> wondering.
>
> Thanks !!!
>
> --
> --
> "Bharat Suneja [MVP]" wrote in message
> news:uUXJ0ZWfIHA.4684@TK2MSFTNGP06.phx.gbl...
>>- This is certainly possible and a common/recommended topology to get high
>>availability locally (CCR) to protect against server and storage/database
>>failure, and get site resilience by replicating to a remote SCR target.
>> - Setting up Storage Groups on a CMS in a clustered environment is no
>> different than setting it up for a standalone/non-clustered server.
>> - The only thing you need to consider (and recommend testing in a lab to
>> be comfortable, before deploying in production), imo, is the path to
>> recovery.
>> (In addition to planning paths for Storage Group + Database on both
>> source and target servers.. they need to be identical).
>> - Understand that SCR target requires activation - there are no automated
>> failures as in the case of CCR/SCC clusters. SCR is not a cluster.
>> - If the SCR source and SCR target are clustered (in this case SCR target
>> should not have a CMS created, just cluster setup and Exchange server
>> setup), you can use /recoverCMS
>> - The SCR target can be a single node cluster (consideration: licensing
>> cost increase).
>>
>> --
>> Bharat Suneja
>> MVP - Exchange
>> www.zenprise.com
>> NEW blog location:
>> exchangepedia.com/blog
>> ----------------------------
>>
>>
>> "Thinkpad21" wrote in message
>> news:uQNk%23NSfIHA.4140@TK2MSFTNGP04.phx.gbl...
>>> Hi,
>>>
>>> For site resiliency design, we are going to SCR to another location the
>>> CCR'd stores on the corporate site.
>>>
>>> I have been looking for exact documentation on how to implement this (
>>> step by step kind of document, not high level design )...
>>>
>>> Wondering what would be the easiest way to configure the Mailbox server
>>> in the resiliency site for SCR ( Standalone or single node cluster ).
>>> This site has about 80 users that will have a local store on this server
>>> as well as having the SCR copy of the corporate site that has about 700
>>> users...
>>>
>>> Since the resiliency site will be very small, we were also planning to
>>> have the HUB/CAS role and Public folders also on this SCR server...
>>>
>>> From what I have read, this is all possible, but wondering where I can
>>> find the implementation steps and also, how to use the SCR copy in case
>>> of site disaster at corporate site, I need to document the steps
>>> required to bring the SCR copy live and have users pointing to it...
>>>
>>> Thanks !!!
>>>
>>> --
>>>
>>>
>>
>
>
date: Mon, 3 Mar 2008 13:19:26 -0800
author: Bharat Suneja [MVP]
Re: SCR of CCR'd stores
That's what I thought, but can I have a store on this server that is not
related to the CCR that is being SCR'd here, or would that also not be
possible, in other words, do I need a dedicated single node cluster for SCR
if I want to use the /RecoverCMS.
Thanks !!!
--
--
---
"Bharat Suneja [MVP]" wrote in message
news:e4DR%23QXfIHA.5296@TK2MSFTNGP05.phx.gbl...
>- /RecoverCMS would be the quickest option, imo.
> - Unfortunately, this means you can't have Hub/CAS on the same box.
> --
> Bharat Suneja
> MVP - Exchange
> www.zenprise.com
> NEW blog location:
> exchangepedia.com/blog
> ----------------------------
>
>
> "Thinkpad21" wrote in message
> news:uiUFJFXfIHA.5280@TK2MSFTNGP02.phx.gbl...
>> Thanks, I guess my question is, will it be easier to activate SCR if the
>> scr target was a single node cluster or stand-alone server...
>>
>> I know that we can use DB portability for the stand-alone scenario, but
>> that means that Outlook profiles need to be modified if Outlook 2007 is
>> not used. So if that's the case, that would require quite an effort...
>>
>> If we go with a SCR that is a single node cluster, than a RecoverCMS
>> would be needed, but in this case, can the server be also a Hub/CAS and
>> have a public folder store ?
>>
>> I have a complete lab infrastructure and test this on it, but was just
>> wondering.
>>
>> Thanks !!!
>>
>> --
>> --
>> "Bharat Suneja [MVP]" wrote in message
>> news:uUXJ0ZWfIHA.4684@TK2MSFTNGP06.phx.gbl...
>>>- This is certainly possible and a common/recommended topology to get
>>>high availability locally (CCR) to protect against server and
>>>storage/database failure, and get site resilience by replicating to a
>>>remote SCR target.
>>> - Setting up Storage Groups on a CMS in a clustered environment is no
>>> different than setting it up for a standalone/non-clustered server.
>>> - The only thing you need to consider (and recommend testing in a lab to
>>> be comfortable, before deploying in production), imo, is the path to
>>> recovery.
>>> (In addition to planning paths for Storage Group + Database on both
>>> source and target servers.. they need to be identical).
>>> - Understand that SCR target requires activation - there are no
>>> automated failures as in the case of CCR/SCC clusters. SCR is not a
>>> cluster.
>>> - If the SCR source and SCR target are clustered (in this case SCR
>>> target should not have a CMS created, just cluster setup and Exchange
>>> server setup), you can use /recoverCMS
>>> - The SCR target can be a single node cluster (consideration: licensing
>>> cost increase).
>>>
>>> --
>>> Bharat Suneja
>>> MVP - Exchange
>>> www.zenprise.com
>>> NEW blog location:
>>> exchangepedia.com/blog
>>> ----------------------------
>>>
>>>
>>> "Thinkpad21" wrote in message
>>> news:uQNk%23NSfIHA.4140@TK2MSFTNGP04.phx.gbl...
>>>> Hi,
>>>>
>>>> For site resiliency design, we are going to SCR to another location the
>>>> CCR'd stores on the corporate site.
>>>>
>>>> I have been looking for exact documentation on how to implement this
>>>> ( step by step kind of document, not high level design )...
>>>>
>>>> Wondering what would be the easiest way to configure the Mailbox server
>>>> in the resiliency site for SCR ( Standalone or single node cluster ).
>>>> This site has about 80 users that will have a local store on this
>>>> server as well as having the SCR copy of the corporate site that has
>>>> about 700 users...
>>>>
>>>> Since the resiliency site will be very small, we were also planning to
>>>> have the HUB/CAS role and Public folders also on this SCR server...
>>>>
>>>> From what I have read, this is all possible, but wondering where I can
>>>> find the implementation steps and also, how to use the SCR copy in case
>>>> of site disaster at corporate site, I need to document the steps
>>>> required to bring the SCR copy live and have users pointing to it...
>>>>
>>>> Thanks !!!
>>>>
>>>> --
>>>>
>>>>
>>>
>>
>>
>
date: Mon, 3 Mar 2008 16:22:42 -0500
author: Thinkpad21
Re: SCR of CCR'd stores
- If you have a single node cluster with no CMS, it won't be able to host
any of its "own" Storage Groups since it doesn't have any identity.
- In other words, a cluster with no CMS isn't really an Exchange server.
There's no msExchExchangeServer object created for it.
- As you've pointed out, you can do this if the target is a
standalone/non-clustered server - it can host its own users on "non-SCRed"
Storage Groups/Stores. In fact, it's own Storage Groups can be replicated to
another SCR target as well. For example, in a 2-site config with standalone
Exchange servers at both locations, ServerA can replicate its SG to ServerB
in remote site. ServerB can do the same with its SG - replicate it to
ServerA in the main/other site.
- The tradeoff is - do you want to lower cost, or lower time to recovery?
--
Bharat Suneja
MVP - Exchange
www.zenprise.com
NEW blog location:
exchangepedia.com/blog
----------------------------
"Thinkpad21" wrote in message
news:%23n4P9TXfIHA.4696@TK2MSFTNGP05.phx.gbl...
> That's what I thought, but can I have a store on this server that is not
> related to the CCR that is being SCR'd here, or would that also not be
> possible, in other words, do I need a dedicated single node cluster for
> SCR if I want to use the /RecoverCMS.
>
> Thanks !!!
>
> --
> --
> ---
> "Bharat Suneja [MVP]" wrote in message
> news:e4DR%23QXfIHA.5296@TK2MSFTNGP05.phx.gbl...
>>- /RecoverCMS would be the quickest option, imo.
>> - Unfortunately, this means you can't have Hub/CAS on the same box.
>> --
>> Bharat Suneja
>> MVP - Exchange
>> www.zenprise.com
>> NEW blog location:
>> exchangepedia.com/blog
>> ----------------------------
>>
>>
>> "Thinkpad21" wrote in message
>> news:uiUFJFXfIHA.5280@TK2MSFTNGP02.phx.gbl...
>>> Thanks, I guess my question is, will it be easier to activate SCR if the
>>> scr target was a single node cluster or stand-alone server...
>>>
>>> I know that we can use DB portability for the stand-alone scenario, but
>>> that means that Outlook profiles need to be modified if Outlook 2007 is
>>> not used. So if that's the case, that would require quite an effort...
>>>
>>> If we go with a SCR that is a single node cluster, than a RecoverCMS
>>> would be needed, but in this case, can the server be also a Hub/CAS and
>>> have a public folder store ?
>>>
>>> I have a complete lab infrastructure and test this on it, but was just
>>> wondering.
>>>
>>> Thanks !!!
>>>
>>> --
>>> --
>>> "Bharat Suneja [MVP]" wrote in message
>>> news:uUXJ0ZWfIHA.4684@TK2MSFTNGP06.phx.gbl...
>>>>- This is certainly possible and a common/recommended topology to get
>>>>high availability locally (CCR) to protect against server and
>>>>storage/database failure, and get site resilience by replicating to a
>>>>remote SCR target.
>>>> - Setting up Storage Groups on a CMS in a clustered environment is no
>>>> different than setting it up for a standalone/non-clustered server.
>>>> - The only thing you need to consider (and recommend testing in a lab
>>>> to be comfortable, before deploying in production), imo, is the path to
>>>> recovery.
>>>> (In addition to planning paths for Storage Group + Database on both
>>>> source and target servers.. they need to be identical).
>>>> - Understand that SCR target requires activation - there are no
>>>> automated failures as in the case of CCR/SCC clusters. SCR is not a
>>>> cluster.
>>>> - If the SCR source and SCR target are clustered (in this case SCR
>>>> target should not have a CMS created, just cluster setup and Exchange
>>>> server setup), you can use /recoverCMS
>>>> - The SCR target can be a single node cluster (consideration: licensing
>>>> cost increase).
>>>>
>>>> --
>>>> Bharat Suneja
>>>> MVP - Exchange
>>>> www.zenprise.com
>>>> NEW blog location:
>>>> exchangepedia.com/blog
>>>> ----------------------------
>>>>
>>>>
>>>> "Thinkpad21" wrote in message
>>>> news:uQNk%23NSfIHA.4140@TK2MSFTNGP04.phx.gbl...
>>>>> Hi,
>>>>>
>>>>> For site resiliency design, we are going to SCR to another location
>>>>> the CCR'd stores on the corporate site.
>>>>>
>>>>> I have been looking for exact documentation on how to implement this
>>>>> ( step by step kind of document, not high level design )...
>>>>>
>>>>> Wondering what would be the easiest way to configure the Mailbox
>>>>> server in the resiliency site for SCR ( Standalone or single node
>>>>> cluster ). This site has about 80 users that will have a local store
>>>>> on this server as well as having the SCR copy of the corporate site
>>>>> that has about 700 users...
>>>>>
>>>>> Since the resiliency site will be very small, we were also planning to
>>>>> have the HUB/CAS role and Public folders also on this SCR server...
>>>>>
>>>>> From what I have read, this is all possible, but wondering where I can
>>>>> find the implementation steps and also, how to use the SCR copy in
>>>>> case of site disaster at corporate site, I need to document the steps
>>>>> required to bring the SCR copy live and have users pointing to it...
>>>>>
>>>>> Thanks !!!
>>>>>
>>>>> --
>>>>>
>>>>>
>>>>
>>>
>>>
>>
>
>
date: Mon, 3 Mar 2008 13:33:21 -0800
author: Bharat Suneja [MVP]
Re: SCR of CCR'd stores
Yep, that's what I thought...
Now if all clients had Outlook 2007, then the database portability would
have been so easy, just tell people to close their outlook and reopen it
after the move of users ( config only ) has been done...and this would work
with a stand-alone SCR that is also housing it's own stores and Hub/cas
Role...
right ???
Thanks for your help !!!
--
"Bharat Suneja [MVP]" wrote in message
news:eJVPwYXfIHA.4140@TK2MSFTNGP04.phx.gbl...
>- If you have a single node cluster with no CMS, it won't be able to host
>any of its "own" Storage Groups since it doesn't have any identity.
> - In other words, a cluster with no CMS isn't really an Exchange server.
> There's no msExchExchangeServer object created for it.
> - As you've pointed out, you can do this if the target is a
> standalone/non-clustered server - it can host its own users on "non-SCRed"
> Storage Groups/Stores. In fact, it's own Storage Groups can be replicated
> to another SCR target as well. For example, in a 2-site config with
> standalone Exchange servers at both locations, ServerA can replicate its
> SG to ServerB in remote site. ServerB can do the same with its SG -
> replicate it to ServerA in the main/other site.
> - The tradeoff is - do you want to lower cost, or lower time to recovery?
> --
> Bharat Suneja
> MVP - Exchange
> www.zenprise.com
> NEW blog location:
> exchangepedia.com/blog
> ----------------------------
>
>
> "Thinkpad21" wrote in message
> news:%23n4P9TXfIHA.4696@TK2MSFTNGP05.phx.gbl...
>> That's what I thought, but can I have a store on this server that is not
>> related to the CCR that is being SCR'd here, or would that also not be
>> possible, in other words, do I need a dedicated single node cluster for
>> SCR if I want to use the /RecoverCMS.
>>
>> Thanks !!!
>>
>> --
>> --
>> ---
>> "Bharat Suneja [MVP]" wrote in message
>> news:e4DR%23QXfIHA.5296@TK2MSFTNGP05.phx.gbl...
>>>- /RecoverCMS would be the quickest option, imo.
>>> - Unfortunately, this means you can't have Hub/CAS on the same box.
>>> --
>>> Bharat Suneja
>>> MVP - Exchange
>>> www.zenprise.com
>>> NEW blog location:
>>> exchangepedia.com/blog
>>> ----------------------------
>>>
>>>
>>> "Thinkpad21" wrote in message
>>> news:uiUFJFXfIHA.5280@TK2MSFTNGP02.phx.gbl...
>>>> Thanks, I guess my question is, will it be easier to activate SCR if
>>>> the scr target was a single node cluster or stand-alone server...
>>>>
>>>> I know that we can use DB portability for the stand-alone scenario, but
>>>> that means that Outlook profiles need to be modified if Outlook 2007 is
>>>> not used. So if that's the case, that would require quite an effort...
>>>>
>>>> If we go with a SCR that is a single node cluster, than a RecoverCMS
>>>> would be needed, but in this case, can the server be also a Hub/CAS and
>>>> have a public folder store ?
>>>>
>>>> I have a complete lab infrastructure and test this on it, but was just
>>>> wondering.
>>>>
>>>> Thanks !!!
>>>>
>>>> --
>>>> --
>>>> "Bharat Suneja [MVP]" wrote in message
>>>> news:uUXJ0ZWfIHA.4684@TK2MSFTNGP06.phx.gbl...
>>>>>- This is certainly possible and a common/recommended topology to get
>>>>>high availability locally (CCR) to protect against server and
>>>>>storage/database failure, and get site resilience by replicating to a
>>>>>remote SCR target.
>>>>> - Setting up Storage Groups on a CMS in a clustered environment is no
>>>>> different than setting it up for a standalone/non-clustered server.
>>>>> - The only thing you need to consider (and recommend testing in a lab
>>>>> to be comfortable, before deploying in production), imo, is the path
>>>>> to recovery.
>>>>> (In addition to planning paths for Storage Group + Database on both
>>>>> source and target servers.. they need to be identical).
>>>>> - Understand that SCR target requires activation - there are no
>>>>> automated failures as in the case of CCR/SCC clusters. SCR is not a
>>>>> cluster.
>>>>> - If the SCR source and SCR target are clustered (in this case SCR
>>>>> target should not have a CMS created, just cluster setup and Exchange
>>>>> server setup), you can use /recoverCMS
>>>>> - The SCR target can be a single node cluster (consideration:
>>>>> licensing cost increase).
>>>>>
>>>>> --
>>>>> Bharat Suneja
>>>>> MVP - Exchange
>>>>> www.zenprise.com
>>>>> NEW blog location:
>>>>> exchangepedia.com/blog
>>>>> ----------------------------
>>>>>
>>>>>
>>>>> "Thinkpad21" wrote in message
>>>>> news:uQNk%23NSfIHA.4140@TK2MSFTNGP04.phx.gbl...
>>>>>> Hi,
>>>>>>
>>>>>> For site resiliency design, we are going to SCR to another location
>>>>>> the CCR'd stores on the corporate site.
>>>>>>
>>>>>> I have been looking for exact documentation on how to implement this
>>>>>> ( step by step kind of document, not high level design )...
>>>>>>
>>>>>> Wondering what would be the easiest way to configure the Mailbox
>>>>>> server in the resiliency site for SCR ( Standalone or single node
>>>>>> cluster ). This site has about 80 users that will have a local store
>>>>>> on this server as well as having the SCR copy of the corporate site
>>>>>> that has about 700 users...
>>>>>>
>>>>>> Since the resiliency site will be very small, we were also planning
>>>>>> to have the HUB/CAS role and Public folders also on this SCR
>>>>>> server...
>>>>>>
>>>>>> From what I have read, this is all possible, but wondering where I
>>>>>> can find the implementation steps and also, how to use the SCR copy
>>>>>> in case of site disaster at corporate site, I need to document the
>>>>>> steps required to bring the SCR copy live and have users pointing to
>>>>>> it...
>>>>>>
>>>>>> Thanks !!!
>>>>>>
>>>>>> --
>>>>>>
>>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>>
>
date: Mon, 3 Mar 2008 17:04:02 -0500
author: Thinkpad21
Re: SCR of CCR'd stores
- That's right - in an all-2007 environment (that is, Exchange and Outlook),
AutoDiscover makes it a snap.
- It is particularly beneficial with standalone boxes (or clustered
implementation where you haven't used Database Portaibility but used
/RecoverCMS instead. In the latter case, the identity of the Exchange Server
remains the same, so you don't need AutoDiscover to make it seamless).
- Having said that, Database Portability comes in handy when the entire
source server hasn't failed - and only a subset of the Storage Groups
replicated using SCR need to be activated.
--
Bharat Suneja
MVP - Exchange
www.zenprise.com
NEW blog location:
exchangepedia.com/blog
----------------------------
"Thinkpad21" wrote in message
news:%23ITX$qXfIHA.5296@TK2MSFTNGP05.phx.gbl...
> Yep, that's what I thought...
>
> Now if all clients had Outlook 2007, then the database portability would
> have been so easy, just tell people to close their outlook and reopen it
> after the move of users ( config only ) has been done...and this would
> work with a stand-alone SCR that is also housing it's own stores and
> Hub/cas Role...
>
> right ???
>
> Thanks for your help !!!
>
>
>
> --
> "Bharat Suneja [MVP]" wrote in message
> news:eJVPwYXfIHA.4140@TK2MSFTNGP04.phx.gbl...
>>- If you have a single node cluster with no CMS, it won't be able to host
>>any of its "own" Storage Groups since it doesn't have any identity.
>> - In other words, a cluster with no CMS isn't really an Exchange server.
>> There's no msExchExchangeServer object created for it.
>> - As you've pointed out, you can do this if the target is a
>> standalone/non-clustered server - it can host its own users on
>> "non-SCRed" Storage Groups/Stores. In fact, it's own Storage Groups can
>> be replicated to another SCR target as well. For example, in a 2-site
>> config with standalone Exchange servers at both locations, ServerA can
>> replicate its SG to ServerB in remote site. ServerB can do the same with
>> its SG - replicate it to ServerA in the main/other site.
>> - The tradeoff is - do you want to lower cost, or lower time to recovery?
>> --
>> Bharat Suneja
>> MVP - Exchange
>> www.zenprise.com
>> NEW blog location:
>> exchangepedia.com/blog
>> ----------------------------
>>
>>
>> "Thinkpad21" wrote in message
>> news:%23n4P9TXfIHA.4696@TK2MSFTNGP05.phx.gbl...
>>> That's what I thought, but can I have a store on this server that is not
>>> related to the CCR that is being SCR'd here, or would that also not be
>>> possible, in other words, do I need a dedicated single node cluster for
>>> SCR if I want to use the /RecoverCMS.
>>>
>>> Thanks !!!
>>>
>>> --
>>> --
>>> ---
>>> "Bharat Suneja [MVP]" wrote in message
>>> news:e4DR%23QXfIHA.5296@TK2MSFTNGP05.phx.gbl...
>>>>- /RecoverCMS would be the quickest option, imo.
>>>> - Unfortunately, this means you can't have Hub/CAS on the same box.
>>>> --
>>>> Bharat Suneja
>>>> MVP - Exchange
>>>> www.zenprise.com
>>>> NEW blog location:
>>>> exchangepedia.com/blog
>>>> ----------------------------
>>>>
>>>>
>>>> "Thinkpad21" wrote in message
>>>> news:uiUFJFXfIHA.5280@TK2MSFTNGP02.phx.gbl...
>>>>> Thanks, I guess my question is, will it be easier to activate SCR if
>>>>> the scr target was a single node cluster or stand-alone server...
>>>>>
>>>>> I know that we can use DB portability for the stand-alone scenario,
>>>>> but that means that Outlook profiles need to be modified if Outlook
>>>>> 2007 is not used. So if that's the case, that would require quite an
>>>>> effort...
>>>>>
>>>>> If we go with a SCR that is a single node cluster, than a RecoverCMS
>>>>> would be needed, but in this case, can the server be also a Hub/CAS
>>>>> and have a public folder store ?
>>>>>
>>>>> I have a complete lab infrastructure and test this on it, but was just
>>>>> wondering.
>>>>>
>>>>> Thanks !!!
>>>>>
>>>>> --
>>>>> --
>>>>> "Bharat Suneja [MVP]" wrote in message
>>>>> news:uUXJ0ZWfIHA.4684@TK2MSFTNGP06.phx.gbl...
>>>>>>- This is certainly possible and a common/recommended topology to get
>>>>>>high availability locally (CCR) to protect against server and
>>>>>>storage/database failure, and get site resilience by replicating to a
>>>>>>remote SCR target.
>>>>>> - Setting up Storage Groups on a CMS in a clustered environment is no
>>>>>> different than setting it up for a standalone/non-clustered server.
>>>>>> - The only thing you need to consider (and recommend testing in a lab
>>>>>> to be comfortable, before deploying in production), imo, is the path
>>>>>> to recovery.
>>>>>> (In addition to planning paths for Storage Group + Database on both
>>>>>> source and target servers.. they need to be identical).
>>>>>> - Understand that SCR target requires activation - there are no
>>>>>> automated failures as in the case of CCR/SCC clusters. SCR is not a
>>>>>> cluster.
>>>>>> - If the SCR source and SCR target are clustered (in this case SCR
>>>>>> target should not have a CMS created, just cluster setup and Exchange
>>>>>> server setup), you can use /recoverCMS
>>>>>> - The SCR target can be a single node cluster (consideration:
>>>>>> licensing cost increase).
>>>>>>
>>>>>> --
>>>>>> Bharat Suneja
>>>>>> MVP - Exchange
>>>>>> www.zenprise.com
>>>>>> NEW blog location:
>>>>>> exchangepedia.com/blog
>>>>>> ----------------------------
>>>>>>
>>>>>>
>>>>>> "Thinkpad21" wrote in message
>>>>>> news:uQNk%23NSfIHA.4140@TK2MSFTNGP04.phx.gbl...
>>>>>>> Hi,
>>>>>>>
>>>>>>> For site resiliency design, we are going to SCR to another location
>>>>>>> the CCR'd stores on the corporate site.
>>>>>>>
>>>>>>> I have been looking for exact documentation on how to implement this
>>>>>>> ( step by step kind of document, not high level design )...
>>>>>>>
>>>>>>> Wondering what would be the easiest way to configure the Mailbox
>>>>>>> server in the resiliency site for SCR ( Standalone or single node
>>>>>>> cluster ). This site has about 80 users that will have a local store
>>>>>>> on this server as well as having the SCR copy of the corporate site
>>>>>>> that has about 700 users...
>>>>>>>
>>>>>>> Since the resiliency site will be very small, we were also planning
>>>>>>> to have the HUB/CAS role and Public folders also on this SCR
>>>>>>> server...
>>>>>>>
>>>>>>> From what I have read, this is all possible, but wondering where I
>>>>>>> can find the implementation steps and also, how to use the SCR copy
>>>>>>> in case of site disaster at corporate site, I need to document the
>>>>>>> steps required to bring the SCR copy live and have users pointing to
>>>>>>> it...
>>>>>>>
>>>>>>> Thanks !!!
>>>>>>>
>>>>>>> --
>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>
>>>
>>
>
>
date: Mon, 3 Mar 2008 21:27:27 -0800
author: Bharat Suneja [MVP]
Re: SCR of CCR'd stores
With database portability in a mixed client environment, you do have to
consider the impact of rehoming a database. If you rehome the entire
server, then a simple cname in DNS will suffice to repoint the clients so
they can pick up the new location of the server on which their mailbox store
resides.
"Bharat Suneja [MVP]" wrote in message
news:eweHyhbfIHA.5900@TK2MSFTNGP02.phx.gbl...
>- That's right - in an all-2007 environment (that is, Exchange and
>Outlook), AutoDiscover makes it a snap.
> - It is particularly beneficial with standalone boxes (or clustered
> implementation where you haven't used Database Portaibility but used
> /RecoverCMS instead. In the latter case, the identity of the Exchange
> Server remains the same, so you don't need AutoDiscover to make it
> seamless).
> - Having said that, Database Portability comes in handy when the entire
> source server hasn't failed - and only a subset of the Storage Groups
> replicated using SCR need to be activated.
> --
> Bharat Suneja
> MVP - Exchange
> www.zenprise.com
> NEW blog location:
> exchangepedia.com/blog
> ----------------------------
>
>
>
> "Thinkpad21" wrote in message
> news:%23ITX$qXfIHA.5296@TK2MSFTNGP05.phx.gbl...
>> Yep, that's what I thought...
>>
>> Now if all clients had Outlook 2007, then the database portability would
>> have been so easy, just tell people to close their outlook and reopen it
>> after the move of users ( config only ) has been done...and this would
>> work with a stand-alone SCR that is also housing it's own stores and
>> Hub/cas Role...
>>
>> right ???
>>
>> Thanks for your help !!!
>>
>>
>>
>> --
>> "Bharat Suneja [MVP]" wrote in message
>> news:eJVPwYXfIHA.4140@TK2MSFTNGP04.phx.gbl...
>>>- If you have a single node cluster with no CMS, it won't be able to host
>>>any of its "own" Storage Groups since it doesn't have any identity.
>>> - In other words, a cluster with no CMS isn't really an Exchange server.
>>> There's no msExchExchangeServer object created for it.
>>> - As you've pointed out, you can do this if the target is a
>>> standalone/non-clustered server - it can host its own users on
>>> "non-SCRed" Storage Groups/Stores. In fact, it's own Storage Groups can
>>> be replicated to another SCR target as well. For example, in a 2-site
>>> config with standalone Exchange servers at both locations, ServerA can
>>> replicate its SG to ServerB in remote site. ServerB can do the same with
>>> its SG - replicate it to ServerA in the main/other site.
>>> - The tradeoff is - do you want to lower cost, or lower time to
>>> recovery?
>>> --
>>> Bharat Suneja
>>> MVP - Exchange
>>> www.zenprise.com
>>> NEW blog location:
>>> exchangepedia.com/blog
>>> ----------------------------
>>>
>>>
>>> "Thinkpad21" wrote in message
>>> news:%23n4P9TXfIHA.4696@TK2MSFTNGP05.phx.gbl...
>>>> That's what I thought, but can I have a store on this server that is
>>>> not related to the CCR that is being SCR'd here, or would that also not
>>>> be possible, in other words, do I need a dedicated single node cluster
>>>> for SCR if I want to use the /RecoverCMS.
>>>>
>>>> Thanks !!!
>>>>
>>>> --
>>>> --
>>>> ---
>>>> "Bharat Suneja [MVP]" wrote in message
>>>> news:e4DR%23QXfIHA.5296@TK2MSFTNGP05.phx.gbl...
>>>>>- /RecoverCMS would be the quickest option, imo.
>>>>> - Unfortunately, this means you can't have Hub/CAS on the same box.
>>>>> --
>>>>> Bharat Suneja
>>>>> MVP - Exchange
>>>>> www.zenprise.com
>>>>> NEW blog location:
>>>>> exchangepedia.com/blog
>>>>> ----------------------------
>>>>>
>>>>>
>>>>> "Thinkpad21" wrote in message
>>>>> news:uiUFJFXfIHA.5280@TK2MSFTNGP02.phx.gbl...
>>>>>> Thanks, I guess my question is, will it be easier to activate SCR if
>>>>>> the scr target was a single node cluster or stand-alone server...
>>>>>>
>>>>>> I know that we can use DB portability for the stand-alone scenario,
>>>>>> but that means that Outlook profiles need to be modified if Outlook
>>>>>> 2007 is not used. So if that's the case, that would require quite an
>>>>>> effort...
>>>>>>
>>>>>> If we go with a SCR that is a single node cluster, than a RecoverCMS
>>>>>> would be needed, but in this case, can the server be also a Hub/CAS
>>>>>> and have a public folder store ?
>>>>>>
>>>>>> I have a complete lab infrastructure and test this on it, but was
>>>>>> just wondering.
>>>>>>
>>>>>> Thanks !!!
>>>>>>
>>>>>> --
>>>>>> --
>>>>>> "Bharat Suneja [MVP]" wrote in message
>>>>>> news:uUXJ0ZWfIHA.4684@TK2MSFTNGP06.phx.gbl...
>>>>>>>- This is certainly possible and a common/recommended topology to get
>>>>>>>high availability locally (CCR) to protect against server and
>>>>>>>storage/database failure, and get site resilience by replicating to a
>>>>>>>remote SCR target.
>>>>>>> - Setting up Storage Groups on a CMS in a clustered environment is
>>>>>>> no different than setting it up for a standalone/non-clustered
>>>>>>> server.
>>>>>>> - The only thing you need to consider (and recommend testing in a
>>>>>>> lab to be comfortable, before deploying in production), imo, is the
>>>>>>> path to recovery.
>>>>>>> (In addition to planning paths for Storage Group + Database on both
>>>>>>> source and target servers.. they need to be identical).
>>>>>>> - Understand that SCR target requires activation - there are no
>>>>>>> automated failures as in the case of CCR/SCC clusters. SCR is not a
>>>>>>> cluster.
>>>>>>> - If the SCR source and SCR target are clustered (in this case SCR
>>>>>>> target should not have a CMS created, just cluster setup and
>>>>>>> Exchange server setup), you can use /recoverCMS
>>>>>>> - The SCR target can be a single node cluster (consideration:
>>>>>>> licensing cost increase).
>>>>>>>
>>>>>>> --
>>>>>>> Bharat Suneja
>>>>>>> MVP - Exchange
>>>>>>> www.zenprise.com
>>>>>>> NEW blog location:
>>>>>>> exchangepedia.com/blog
>>>>>>> ----------------------------
>>>>>>>
>>>>>>>
>>>>>>> "Thinkpad21" wrote in message
>>>>>>> news:uQNk%23NSfIHA.4140@TK2MSFTNGP04.phx.gbl...
>>>>>>>> Hi,
>>>>>>>>
>>>>>>>> For site resiliency design, we are going to SCR to another location
>>>>>>>> the CCR'd stores on the corporate site.
>>>>>>>>
>>>>>>>> I have been looking for exact documentation on how to implement
>>>>>>>> this ( step by step kind of document, not high level design )...
>>>>>>>>
>>>>>>>> Wondering what would be the easiest way to configure the Mailbox
>>>>>>>> server in the resiliency site for SCR ( Standalone or single node
>>>>>>>> cluster ). This site has about 80 users that will have a local
>>>>>>>> store on this server as well as having the SCR copy of the
>>>>>>>> corporate site that has about 700 users...
>>>>>>>>
>>>>>>>> Since the resiliency site will be very small, we were also planning
>>>>>>>> to have the HUB/CAS role and Public folders also on this SCR
>>>>>>>> server...
>>>>>>>>
>>>>>>>> From what I have read, this is all possible, but wondering where I
>>>>>>>> can find the implementation steps and also, how to use the SCR copy
>>>>>>>> in case of site disaster at corporate site, I need to document the
>>>>>>>> steps required to bring the SCR copy live and have users pointing
>>>>>>>> to it...
>>>>>>>>
>>>>>>>> Thanks !!!
>>>>>>>>
>>>>>>>> --
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>>
>
date: Tue, 4 Mar 2008 19:36:56 -0800
author: John Fullbright fjohn@donotspamnetappdotcom
|
|
