RE: Questions about DAG's (Database Availability Groups) in exchange 2010
Hi Thank you for all your replies. This has given me a better understanding of what is expected to happen. One more question though. (For the moment) When an exchange server fails and all the clients reconnect to another exchange server. What happens when the original exchange server becomes available again? Do the Outlook clients automatically change back to using their original Exchange server? Or do they just continue to connect to the server that they have been connected to whilst their (home) server was off line? I am just thinking about our circumstances where we have one exchange server in the UK and one In the US. It would seem illogical for the clients to continue to connect to a server in the US if the one in our office became available again after a failure. And vice versa should a failure occur on the US server. Regards Kevan Dickinson Network Manager NSF-CMI 23 Lodge Road Hanborough Business Park, Long Hanborough, Oxford, OX29 8SJ, UK T:+44 01993 885661 E:kevan.dickin...@nsf-cmi.com W:www.nsf-cmi.com From: Michael B. Smith [mailto:mich...@smithcons.com] Sent: 07 April 2011 01:11 To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 In order to answer this question completely (and accurately) depends on a full understanding of your topology and Exchange deployment. The best I can say to some of your questions, based on what you've told us, is: it depends. :) Failover within a DAG happens automatically. Given healthy replication and 80ms or less latency between the servers, you should see failover within 30 seconds. FailBACK (which is actually referred to as a switchover) is a manual process. Insofar as how communications happens with HTs and CAS - insufficient data. Having a GLBS makes some things easier, but if you have redundancy designed into your topology, it isn't necessary - but it depends on what you are protecting and how you've done your implementation. There is a great deal of flexibility that allows you to design to meet the specific needs of your company and the behavior is dependent on that design. Regards, Michael B. Smith Consultant and Exchange MVP http://TheEssentialExchange.com From: Kevan Dickinson [mailto:kevan.dickin...@nsf-cmi.com] Sent: Wednesday, April 06, 2011 6:09 PM To: MS-Exchange Admin Issues Subject: Questions about DAG's (Database Availability Groups) in exchange 2010 All. This is our Current setup. Exchange 2010 servers in the UK and the US both in the same Exchange organization. At the moment we are running a Windows 2003 Domain structure. The UK domain is a child domain of the US Domain. I was wondering if someone could answer a question for me about DAG's in Exchange 2010. It is something I would like to understand in case we would like to implement it. My understanding of a DAG is that it is a replication of the Mailbox database attached to an Exchange 2010 server to another site / server in case the main database becomes corrupt. However what happens If the actual server that the mailbox database is attached to becomes unavailable will the users who have mailboxes on the unavailable server be automatically diverted to another server where the replicated database is? Or would you need to attach the replicated database to another server manually in order for users to become attached to their email again? What would happen in a situation like ours if say we replicated our Database to our office in the US and our server became unavailable. Would the office in the US need to manually mount our database on their server and then all our external and internal users need to change their outlook settings to look for client.USCompany.org? Instead of client.ukcompany.com If so presumably there mail would then get routed via there anti virus / spam filtering software? I am trying to work out what is the best way to get better exchange redundancy in the event of either the An US Exchange 2010 server or ours being becoming unavailable? Presumably everything would be much easier if we had one Exchange Gateway, or does this not matter really? At the moment email in the UK enters via our Mail gateway / anti virus and anti spam system and email in the US enters via there gateway. You help and comments would be appreciated. Regards Kevan ***Disclaimer*** The contents of this Email may be privileged and are confidential. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited and may be unlawful. Should you wish to use Email as a mode of communication, NSF-CMi Ltd and its subsidiaries are unable to guarantee the security of Email content outside of our own computer systems. This footnote also confirms that this Email message has been checked
RE: Questions about DAG's (Database Availability Groups) in exchange 2010
You are thinking in pre Exchange 2010 terms. A given database can only be on-line on one server at any given point in time, so clients always access the e-mail on the server with the database on-line. However if I understand things properly, clients now never access the database directly. They always access via the Client Access role. So if the current client access server role remains active they will continue to connect to that client access server. Its role is to figure out where the mail box is and manage access to it in a transparent manner. Dave Wade From: Kevan Dickinson [mailto:kevan.dickin...@nsf-cmi.com] Sent: 07 April 2011 10:14 To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 Hi Thank you for all your replies. This has given me a better understanding of what is expected to happen. One more question though. (For the moment) When an exchange server fails and all the clients reconnect to another exchange server. What happens when the original exchange server becomes available again? Do the Outlook clients automatically change back to using their original Exchange server? Or do they just continue to connect to the server that they have been connected to whilst their (home) server was off line? I am just thinking about our circumstances where we have one exchange server in the UK and one In the US. It would seem illogical for the clients to continue to connect to a server in the US if the one in our office became available again after a failure. And vice versa should a failure occur on the US server. Regards Kevan Dickinson Network Manager NSF-CMI 23 Lodge Road Hanborough Business Park, Long Hanborough, Oxford, OX29 8SJ, UK T:+44 01993 885661 E:kevan.dickin...@nsf-cmi.com W:www.nsf-cmi.com From: Michael B. Smith [mailto:mich...@smithcons.com] Sent: 07 April 2011 01:11 To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 In order to answer this question completely (and accurately) depends on a full understanding of your topology and Exchange deployment. The best I can say to some of your questions, based on what you've told us, is: it depends. J Failover within a DAG happens automatically. Given healthy replication and 80ms or less latency between the servers, you should see failover within 30 seconds. FailBACK (which is actually referred to as a switchover) is a manual process. Insofar as how communications happens with HTs and CAS - insufficient data. Having a GLBS makes some things easier, but if you have redundancy designed into your topology, it isn't necessary - but it depends on what you are protecting and how you've done your implementation. There is a great deal of flexibility that allows you to design to meet the specific needs of your company and the behavior is dependent on that design. Regards, Michael B. Smith Consultant and Exchange MVP http://TheEssentialExchange.com From: Kevan Dickinson [mailto:kevan.dickin...@nsf-cmi.com] Sent: Wednesday, April 06, 2011 6:09 PM To: MS-Exchange Admin Issues Subject: Questions about DAG's (Database Availability Groups) in exchange 2010 All. This is our Current setup. Exchange 2010 servers in the UK and the US both in the same Exchange organization. At the moment we are running a Windows 2003 Domain structure. The UK domain is a child domain of the US Domain. I was wondering if someone could answer a question for me about DAG's in Exchange 2010. It is something I would like to understand in case we would like to implement it. My understanding of a DAG is that it is a replication of the Mailbox database attached to an Exchange 2010 server to another site / server in case the main database becomes corrupt. However what happens If the actual server that the mailbox database is attached to becomes unavailable will the users who have mailboxes on the unavailable server be automatically diverted to another server where the replicated database is? Or would you need to attach the replicated database to another server manually in order for users to become attached to their email again? What would happen in a situation like ours if say we replicated our Database to our office in the US and our server became unavailable. Would the office in the US need to manually mount our database
RE: Questions about DAG's (Database Availability Groups) in exchange 2010
FailBACK (which is actually referred to as a switchover) is a manual process. That applies to mailbox servers. HT/CAS - depends on your topology. Generally, you don't break existing connections except by manual effort. Regards, Michael B. Smith Consultant and Exchange MVP http://TheEssentialExchange.com From: Kevan Dickinson [mailto:kevan.dickin...@nsf-cmi.com] Sent: Thursday, April 07, 2011 5:14 AM To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 Hi Thank you for all your replies. This has given me a better understanding of what is expected to happen. One more question though. (For the moment) When an exchange server fails and all the clients reconnect to another exchange server. What happens when the original exchange server becomes available again? Do the Outlook clients automatically change back to using their original Exchange server? Or do they just continue to connect to the server that they have been connected to whilst their (home) server was off line? I am just thinking about our circumstances where we have one exchange server in the UK and one In the US. It would seem illogical for the clients to continue to connect to a server in the US if the one in our office became available again after a failure. And vice versa should a failure occur on the US server. Regards Kevan Dickinson Network Manager NSF-CMI 23 Lodge Road Hanborough Business Park, Long Hanborough, Oxford, OX29 8SJ, UK T:+44 01993 885661 E:kevan.dickin...@nsf-cmi.com W:www.nsf-cmi.com From: Michael B. Smith [mailto:mich...@smithcons.com] Sent: 07 April 2011 01:11 To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 In order to answer this question completely (and accurately) depends on a full understanding of your topology and Exchange deployment. The best I can say to some of your questions, based on what you've told us, is: it depends. :) Failover within a DAG happens automatically. Given healthy replication and 80ms or less latency between the servers, you should see failover within 30 seconds. FailBACK (which is actually referred to as a switchover) is a manual process. Insofar as how communications happens with HTs and CAS - insufficient data. Having a GLBS makes some things easier, but if you have redundancy designed into your topology, it isn't necessary - but it depends on what you are protecting and how you've done your implementation. There is a great deal of flexibility that allows you to design to meet the specific needs of your company and the behavior is dependent on that design. Regards, Michael B. Smith Consultant and Exchange MVP http://TheEssentialExchange.com From: Kevan Dickinson [mailto:kevan.dickin...@nsf-cmi.com] Sent: Wednesday, April 06, 2011 6:09 PM To: MS-Exchange Admin Issues Subject: Questions about DAG's (Database Availability Groups) in exchange 2010 All. This is our Current setup. Exchange 2010 servers in the UK and the US both in the same Exchange organization. At the moment we are running a Windows 2003 Domain structure. The UK domain is a child domain of the US Domain. I was wondering if someone could answer a question for me about DAG's in Exchange 2010. It is something I would like to understand in case we would like to implement it. My understanding of a DAG is that it is a replication of the Mailbox database attached to an Exchange 2010 server to another site / server in case the main database becomes corrupt. However what happens If the actual server that the mailbox database is attached to becomes unavailable will the users who have mailboxes on the unavailable server be automatically diverted to another server where the replicated database is? Or would you need to attach the replicated database to another server manually in order for users to become attached to their email again? What would happen in a situation like ours if say we replicated our Database to our office in the US and our server became unavailable. Would the office in the US need to manually mount our database on their server and then all our external and internal users need to change their outlook settings to look for client.USCompany.org? Instead of client.ukcompany.com If so presumably there mail would then get routed via there anti virus / spam filtering software? I am trying to work out what is the best way to get better exchange redundancy in the event of either the An US Exchange 2010 server or ours being becoming unavailable? Presumably everything would be much easier if we had one Exchange Gateway, or does this not matter really? At the moment email in the UK enters via our Mail gateway / anti virus and anti spam system and email in the US enters via there gateway. You help and comments would be appreciated. Regards Kevan
RE: Questions about DAG's (Database Availability Groups) in exchange 2010
Hi All How do you cope with the scenario for external users, where by access to point A webmail.companyname.org in the US is unavailable and Outlook anywhere users need to connect instead to Access Point B owa.companyname.com in the UK. Apart from manually updating DNS entries for webmail.companyname.org to point at access point B. is there some way of setting up DNS or some kind of failover to automate this process? We are running Outlook versions 2007 and 2010. The DNS entries for Autodiscover would presumably need to be updated as well to point at the temporary access point unless this can be automated in some way. regards Kevan Dickinson Network Manager NSF-CMI 23 Lodge Road Hanborough Business Park, Long Hanborough, Oxford, OX29 8SJ, UK T:+44 01993 885661 E:kevan.dickin...@nsf-cmi.com W:www.nsf-cmi.com From: Michael B. Smith [mailto:mich...@smithcons.com] Sent: 07 April 2011 12:25 To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 FailBACK (which is actually referred to as a switchover) is a manual process. That applies to mailbox servers. HT/CAS - depends on your topology. Generally, you don't break existing connections except by manual effort. Regards, Michael B. Smith Consultant and Exchange MVP http://TheEssentialExchange.com From: Kevan Dickinson [mailto:kevan.dickin...@nsf-cmi.com] Sent: Thursday, April 07, 2011 5:14 AM To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 Hi Thank you for all your replies. This has given me a better understanding of what is expected to happen. One more question though. (For the moment) When an exchange server fails and all the clients reconnect to another exchange server. What happens when the original exchange server becomes available again? Do the Outlook clients automatically change back to using their original Exchange server? Or do they just continue to connect to the server that they have been connected to whilst their (home) server was off line? I am just thinking about our circumstances where we have one exchange server in the UK and one In the US. It would seem illogical for the clients to continue to connect to a server in the US if the one in our office became available again after a failure. And vice versa should a failure occur on the US server. Regards Kevan Dickinson Network Manager NSF-CMI 23 Lodge Road Hanborough Business Park, Long Hanborough, Oxford, OX29 8SJ, UK T:+44 01993 885661 E:kevan.dickin...@nsf-cmi.com W:www.nsf-cmi.com From: Michael B. Smith [mailto:mich...@smithcons.com] Sent: 07 April 2011 01:11 To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 In order to answer this question completely (and accurately) depends on a full understanding of your topology and Exchange deployment. The best I can say to some of your questions, based on what you've told us, is: it depends. :) Failover within a DAG happens automatically. Given healthy replication and 80ms or less latency between the servers, you should see failover within 30 seconds. FailBACK (which is actually referred to as a switchover) is a manual process. Insofar as how communications happens with HTs and CAS - insufficient data. Having a GLBS makes some things easier, but if you have redundancy designed into your topology, it isn't necessary - but it depends on what you are protecting and how you've done your implementation. There is a great deal of flexibility that allows you to design to meet the specific needs of your company and the behavior is dependent on that design. Regards, Michael B. Smith Consultant and Exchange MVP http://TheEssentialExchange.com From: Kevan Dickinson [mailto:kevan.dickin...@nsf-cmi.com] Sent: Wednesday, April 06, 2011 6:09 PM To: MS-Exchange Admin Issues Subject: Questions about DAG's (Database Availability Groups) in exchange 2010 All. This is our Current setup. Exchange 2010 servers in the UK and the US both in the same Exchange organization. At the moment we are running a Windows 2003 Domain structure. The UK domain is a child domain of the US Domain. I was wondering if someone could answer a question for me about DAG's in Exchange 2010. It is something I would like to understand in case we would like to implement it. My understanding of a DAG is that it is a replication of the Mailbox database attached to an Exchange 2010 server to another site / server in case the main database becomes corrupt. However what happens If the actual server that the mailbox database is attached to becomes unavailable will the users who have mailboxes on the unavailable server be automatically diverted to another server where the replicated database is? Or would you need to attach the replicated database to another server manually in order for users to become attached to their email again? What would
RE: Questions about DAG's (Database Availability Groups) in exchange 2010
Mr. Dickinson, This article is a good source for installation and it may address some of your questions about basic operation; most of which have been answered by others in this message thread. Exchange 2010 RTM DAG using Server 2008 R2 - Part 1http://www.shudnow.net/2009/10/29/exchange-2010-rtm-dag-using-server-2008-r2-%E2%80%93-part-1/ There are four parts Please note that this article is referring to Exchange RTM and Server 2008 R2 pre any service packs and there are some changes with these SP's. I have found this a good written reference reading material on installation that extends a bit beyond this. With regard to your question below, sounding a bit obtuse, it depends on what your goal is. If you wish to have an immediate for a total failure in the US; this can be a challenge as there will be a lag time while DNS entries are updated outside your organization. An alternative solution would be to utilize OWA for temporary outages and instruct users to redirect to another site via separate URL's. If there is a partial failure (of database servers) then there are ways to utilize front end servers (which I recommend) like TMG or F5 to handle redirection to separate CAS arrays. Getting information from sources like this is a good first step. Partnering with other professionals (both formal and informal) is the next logical choice. Planning for failover scenarios in a multi site organization, while not extremely complicated, is a process that starts with clearly outlined goals. I cannot stress enough the importance of having a clear objective prior to architecting a solution. Kindest regards, Steven Alfano Sr. Systems Administrator The Rockefeller University 1230 York Avenue New York, NY 10065-6399 Voice 212.327.8937 Mobile 646.438.5160 fax 212.327.8712 salf...@rockefeller.edu www.rockefeller.edu From: Kevan Dickinson [mailto:kevan.dickin...@nsf-cmi.com] Sent: Thursday, April 07, 2011 8:47 AM To: MS-Exchange Admin Issues Subject: **PHISHING?** RE: Questions about DAG's (Database Availability Groups) in exchange 2010 Hi All How do you cope with the scenario for external users, where by access to point A webmail.companyname.org in the US is unavailable and Outlook anywhere users need to connect instead to Access Point B owa.companyname.com in the UK. Apart from manually updating DNS entries for webmail.companyname.org to point at access point B. is there some way of setting up DNS or some kind of failover to automate this process? We are running Outlook versions 2007 and 2010. The DNS entries for Autodiscover would presumably need to be updated as well to point at the temporary access point unless this can be automated in some way. regards Kevan Dickinson Network Manager NSF-CMI 23 Lodge Road Hanborough Business Park, Long Hanborough, Oxford, OX29 8SJ, UK T:+44 01993 885661 E:kevan.dickin...@nsf-cmi.com W:www.nsf-cmi.com From: Michael B. Smith [mailto:mich...@smithcons.com] Sent: 07 April 2011 12:25 To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 FailBACK (which is actually referred to as a switchover) is a manual process. That applies to mailbox servers. HT/CAS - depends on your topology. Generally, you don't break existing connections except by manual effort. Regards, Michael B. Smith Consultant and Exchange MVP http://TheEssentialExchange.com From: Kevan Dickinson [mailto:kevan.dickin...@nsf-cmi.com] Sent: Thursday, April 07, 2011 5:14 AM To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 Hi Thank you for all your replies. This has given me a better understanding of what is expected to happen. One more question though. (For the moment) When an exchange server fails and all the clients reconnect to another exchange server. What happens when the original exchange server becomes available again? Do the Outlook clients automatically change back to using their original Exchange server? Or do they just continue to connect to the server that they have been connected to whilst their (home) server was off line? I am just thinking about our circumstances where we have one exchange server in the UK and one In the US. It would seem illogical for the clients to continue to connect to a server in the US if the one in our office became available again after a failure. And vice versa should a failure occur on the US server. Regards Kevan Dickinson Network Manager NSF-CMI 23 Lodge Road Hanborough Business Park, Long Hanborough, Oxford, OX29 8SJ, UK T:+44 01993 885661 E:kevan.dickin...@nsf-cmi.com W:www.nsf-cmi.com From: Michael B. Smith [mailto:mich...@smithcons.com] Sent: 07 April 2011 01:11 To: MS-Exchange Admin Issues Subject: RE: Questions about DAG's (Database Availability Groups) in exchange 2010 In order to answer this question completely (and accurately) depends on a full understanding of your
RE: Questions about DAG's (Database Availability Groups) in exchange 2010
I can't answer your cross site/continent question, but a database in a DAG is designed to failover when a database becomes unavailable for any reason, whether the hardware has failed, database access issue, or someone simple rebooted a server by accident :) Mail delivery continues as normal, Outlook users experience a brief period of disconnected state whilst it fails over and the CAS switches the client to the new database. Richard From: bounce-9312473-8066...@lyris.sunbelt-software.com [mailto:bounce-9312473-8066...@lyris.sunbelt-software.com] On Behalf Of Kevan Dickinson Sent: 06 April 2011 23:09 To: MS-Exchange Admin Issues Subject: Questions about DAG's (Database Availability Groups) in exchange 2010 All. This is our Current setup. Exchange 2010 servers in the UK and the US both in the same Exchange organization. At the moment we are running a Windows 2003 Domain structure. The UK domain is a child domain of the US Domain. I was wondering if someone could answer a question for me about DAG's in Exchange 2010. It is something I would like to understand in case we would like to implement it. My understanding of a DAG is that it is a replication of the Mailbox database attached to an Exchange 2010 server to another site / server in case the main database becomes corrupt. However what happens If the actual server that the mailbox database is attached to becomes unavailable will the users who have mailboxes on the unavailable server be automatically diverted to another server where the replicated database is? Or would you need to attach the replicated database to another server manually in order for users to become attached to their email again? What would happen in a situation like ours if say we replicated our Database to our office in the US and our server became unavailable. Would the office in the US need to manually mount our database on their server and then all our external and internal users need to change their outlook settings to look for client.USCompany.org? Instead of client.ukcompany.com If so presumably there mail would then get routed via there anti virus / spam filtering software? I am trying to work out what is the best way to get better exchange redundancy in the event of either the An US Exchange 2010 server or ours being becoming unavailable? Presumably everything would be much easier if we had one Exchange Gateway, or does this not matter really? At the moment email in the UK enters via our Mail gateway / anti virus and anti spam system and email in the US enters via there gateway. You help and comments would be appreciated. Regards Kevan ***Disclaimer*** The contents of this Email may be privileged and are confidential. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited and may be unlawful. Should you wish to use Email as a mode of communication, NSF-CMi Ltd and its subsidiaries are unable to guarantee the security of Email content outside of our own computer systems. This footnote also confirms that this Email message has been checked by MailMarshal for the presence of computer viruses. Whilst we run anti-virus software, you are solely responsible for ensuring that any Email or attachment you receive is virus free. We disclaim any liability for any damage you suffer as a consequence of receiving any virus. NSF-CMi Ltd Registered in England No: 1899857 Registered Office : Hill House, 1 Little New Street, London, EC4A 3TR. ** --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ or send an email to listmana...@lyris.sunbeltsoftware.commailto:listmana...@lyris.sunbeltsoftware.com with the body: unsubscribe exchangelist --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ or send an email to listmana...@lyris.sunbeltsoftware.com with the body: unsubscribe exchangelist
Re: Questions about DAG's (Database Availability Groups) in exchange 2010
30 to 45 seconds to see the failure and a few seconds to do an automated failover. Derek A Johnson Sr. Systems Administrator National Association of Realtors 430 N. Michigan Ave. Chicago, IL 60611 Email: mailto:djohn...@realtors.org djohn...@realtors.orgmailto:djohn...@realtors.org Cell: 262.496.9201 Desk: 312.329.8618 http://www.realtors.org Sent from my iPhone. Please excuse what auto correct can't catch or has auto not corrected. On Apr 6, 2011, at 17:38, Sobey, Richard A r.so...@imperial.ac.ukmailto:r.so...@imperial.ac.uk wrote: I can’t answer your cross site/continent question, but a database in a DAG is designed to failover when a database becomes unavailable for any reason, whether the hardware has failed, database access issue, or someone simple rebooted a server by accident :) Mail delivery continues as normal, Outlook users experience a brief period of disconnected state whilst it fails over and the CAS switches the client to the new database. Richard From: bounce-9312473-8066...@lyris.sunbelt-software.commailto:bounce-9312473-8066...@lyris.sunbelt-software.com [mailto:bounce-9312473-8066...@lyris.sunbelt-software.com] On Behalf Of Kevan Dickinson Sent: 06 April 2011 23:09 To: MS-Exchange Admin Issues Subject: Questions about DAG's (Database Availability Groups) in exchange 2010 All. This is our Current setup. Exchange 2010 servers in the UK and the US both in the same Exchange organization. At the moment we are running a Windows 2003 Domain structure. The UK domain is a child domain of the US Domain. I was wondering if someone could answer a question for me about DAG’s in Exchange 2010. It is something I would like to understand in case we would like to implement it. My understanding of a DAG is that it is a replication of the Mailbox database attached to an Exchange 2010 server to another site / server in case the main database becomes corrupt. However what happens If the actual server that the mailbox database is attached to becomes unavailable will the users who have mailboxes on the unavailable server be automatically diverted to another server where the replicated database is? Or would you need to attach the replicated database to another server manually in order for users to become attached to their email again? What would happen in a situation like ours if say we replicated our Database to our office in the US and our server became unavailable. Would the office in the US need to manually mount our database on their server and then all our external and internal users need to change their outlook settings to look for client.USCompany.orghttp://client.USCompany.org? Instead of client.ukcompany.comhttp://client.ukcompany.com If so presumably there mail would then get routed via there anti virus / spam filtering software? I am trying to work out what is the best way to get better exchange redundancy in the event of either the An US Exchange 2010 server or ours being becoming unavailable? Presumably everything would be much easier if we had one Exchange Gateway, or does this not matter really? At the moment email in the UK enters via our Mail gateway / anti virus and anti spam system and email in the US enters via there gateway. You help and comments would be appreciated. Regards Kevan ***Disclaimer*** The contents of this Email may be privileged and are confidential. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited and may be unlawful. Should you wish to use Email as a mode of communication, NSF-CMi Ltd and its subsidiaries are unable to guarantee the security of Email content outside of our own computer systems. This footnote also confirms that this Email message has been checked by MailMarshal for the presence of computer viruses. Whilst we run anti-virus software, you are solely responsible for ensuring that any Email or attachment you receive is virus free. We disclaim any liability for any damage you suffer as a consequence of receiving any virus. NSF-CMi Ltd Registered in England No: 1899857 Registered Office : Hill House, 1 Little New Street, London, EC4A 3TR. ** --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ http://lyris.sunbelt-software.com/read/my_forums/ or send an email to mailto:listmana...@lyris.sunbeltsoftware.com listmana...@lyris.sunbeltsoftware.commailto:listmana...@lyris.sunbeltsoftware.com with the body: unsubscribe exchangelist --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ http://lyris.sunbelt-software.com/read/my_forums/ or send an email to mailto:listmana...@lyris.sunbeltsoftware.com
RE: Questions about DAG's (Database Availability Groups) in exchange 2010
Less than 30 seconds is the documented target (and what I see when I've tested it - but that depends on a number of items). Regards, Michael B. Smith Consultant and Exchange MVP http://TheEssentialExchange.com From: Derek Johnson [mailto:djohn...@realtors.org] Sent: Wednesday, April 06, 2011 6:42 PM To: MS-Exchange Admin Issues Subject: Re: Questions about DAG's (Database Availability Groups) in exchange 2010 30 to 45 seconds to see the failure and a few seconds to do an automated failover. Derek A Johnson Sr. Systems Administrator National Association of Realtors 430 N. Michigan Ave. Chicago, IL 60611 Email: djohn...@realtors.orgmailto:djohn...@realtors.org Cell: 262.496.9201 Desk: 312.329.8618 http://www.realtors.org Sent from my iPhone. Please excuse what auto correct can't catch or has auto not corrected. On Apr 6, 2011, at 17:38, Sobey, Richard A r.so...@imperial.ac.ukmailto:r.so...@imperial.ac.uk wrote: I can't answer your cross site/continent question, but a database in a DAG is designed to failover when a database becomes unavailable for any reason, whether the hardware has failed, database access issue, or someone simple rebooted a server by accident :) Mail delivery continues as normal, Outlook users experience a brief period of disconnected state whilst it fails over and the CAS switches the client to the new database. Richard From: bounce-9312473-8066...@lyris.sunbelt-software.commailto:bounce-9312473-8066...@lyris.sunbelt-software.com [mailto:bounce-9312473-8066...@lyris.sunbelt-software.com] On Behalf Of Kevan Dickinson Sent: 06 April 2011 23:09 To: MS-Exchange Admin Issues Subject: Questions about DAG's (Database Availability Groups) in exchange 2010 All. This is our Current setup. Exchange 2010 servers in the UK and the US both in the same Exchange organization. At the moment we are running a Windows 2003 Domain structure. The UK domain is a child domain of the US Domain. I was wondering if someone could answer a question for me about DAG's in Exchange 2010. It is something I would like to understand in case we would like to implement it. My understanding of a DAG is that it is a replication of the Mailbox database attached to an Exchange 2010 server to another site / server in case the main database becomes corrupt. However what happens If the actual server that the mailbox database is attached to becomes unavailable will the users who have mailboxes on the unavailable server be automatically diverted to another server where the replicated database is? Or would you need to attach the replicated database to another server manually in order for users to become attached to their email again? What would happen in a situation like ours if say we replicated our Database to our office in the US and our server became unavailable. Would the office in the US need to manually mount our database on their server and then all our external and internal users need to change their outlook settings to look for client.USCompany.orghttp://client.USCompany.org? Instead of client.ukcompany.comhttp://client.ukcompany.com If so presumably there mail would then get routed via there anti virus / spam filtering software? I am trying to work out what is the best way to get better exchange redundancy in the event of either the An US Exchange 2010 server or ours being becoming unavailable? Presumably everything would be much easier if we had one Exchange Gateway, or does this not matter really? At the moment email in the UK enters via our Mail gateway / anti virus and anti spam system and email in the US enters via there gateway. You help and comments would be appreciated. Regards Kevan ***Disclaimer*** The contents of this Email may be privileged and are confidential. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited and may be unlawful. Should you wish to use Email as a mode of communication, NSF-CMi Ltd and its subsidiaries are unable to guarantee the security of Email content outside of our own computer systems. This footnote also confirms that this Email message has been checked by MailMarshal for the presence of computer viruses. Whilst we run anti-virus software, you are solely responsible for ensuring that any Email or attachment you receive is virus free. We disclaim any liability for any damage you suffer as a consequence of receiving any virus. NSF-CMi Ltd Registered in England No: 1899857 Registered Office : Hill House, 1 Little New Street, London, EC4A 3TR. ** --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ or send an email to listmana...@lyris.sunbeltsoftware.commailto:listmana...@lyris.sunbeltsoftware.com
RE: Questions about DAG's (Database Availability Groups) in exchange 2010
In order to answer this question completely (and accurately) depends on a full understanding of your topology and Exchange deployment. The best I can say to some of your questions, based on what you've told us, is: it depends. :) Failover within a DAG happens automatically. Given healthy replication and 80ms or less latency between the servers, you should see failover within 30 seconds. FailBACK (which is actually referred to as a switchover) is a manual process. Insofar as how communications happens with HTs and CAS - insufficient data. Having a GLBS makes some things easier, but if you have redundancy designed into your topology, it isn't necessary - but it depends on what you are protecting and how you've done your implementation. There is a great deal of flexibility that allows you to design to meet the specific needs of your company and the behavior is dependent on that design. Regards, Michael B. Smith Consultant and Exchange MVP http://TheEssentialExchange.com From: Kevan Dickinson [mailto:kevan.dickin...@nsf-cmi.com] Sent: Wednesday, April 06, 2011 6:09 PM To: MS-Exchange Admin Issues Subject: Questions about DAG's (Database Availability Groups) in exchange 2010 All. This is our Current setup. Exchange 2010 servers in the UK and the US both in the same Exchange organization. At the moment we are running a Windows 2003 Domain structure. The UK domain is a child domain of the US Domain. I was wondering if someone could answer a question for me about DAG's in Exchange 2010. It is something I would like to understand in case we would like to implement it. My understanding of a DAG is that it is a replication of the Mailbox database attached to an Exchange 2010 server to another site / server in case the main database becomes corrupt. However what happens If the actual server that the mailbox database is attached to becomes unavailable will the users who have mailboxes on the unavailable server be automatically diverted to another server where the replicated database is? Or would you need to attach the replicated database to another server manually in order for users to become attached to their email again? What would happen in a situation like ours if say we replicated our Database to our office in the US and our server became unavailable. Would the office in the US need to manually mount our database on their server and then all our external and internal users need to change their outlook settings to look for client.USCompany.org? Instead of client.ukcompany.com If so presumably there mail would then get routed via there anti virus / spam filtering software? I am trying to work out what is the best way to get better exchange redundancy in the event of either the An US Exchange 2010 server or ours being becoming unavailable? Presumably everything would be much easier if we had one Exchange Gateway, or does this not matter really? At the moment email in the UK enters via our Mail gateway / anti virus and anti spam system and email in the US enters via there gateway. You help and comments would be appreciated. Regards Kevan ***Disclaimer*** The contents of this Email may be privileged and are confidential. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited and may be unlawful. Should you wish to use Email as a mode of communication, NSF-CMi Ltd and its subsidiaries are unable to guarantee the security of Email content outside of our own computer systems. This footnote also confirms that this Email message has been checked by MailMarshal for the presence of computer viruses. Whilst we run anti-virus software, you are solely responsible for ensuring that any Email or attachment you receive is virus free. We disclaim any liability for any damage you suffer as a consequence of receiving any virus. NSF-CMi Ltd Registered in England No: 1899857 Registered Office : Hill House, 1 Little New Street, London, EC4A 3TR. ** --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ or send an email to listmana...@lyris.sunbeltsoftware.commailto:listmana...@lyris.sunbeltsoftware.com with the body: unsubscribe exchangelist --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ or send an email to listmana...@lyris.sunbeltsoftware.com with the body: unsubscribe exchangelist
