|
|
You're in the
sql server clustering
group:SAN drive config for SQL Cluster
sql server clustering:
What's the best physical drive configuration for a SQL cluster on a SAN?
Would creating a large (14drive/2buses) physcial RAID 1+0 array with appropriate sized LUNs be acceptable for the quorum, tempdb, ms-dtc, and sql tlog shared resources? Or would individual mirrored drives defined as LUNs be better? Thanks,
With SQL 2000 you don't separate tempdb anymore.
If you use MSDTC that should be on it's own disk - if you don't use it for any applications - don't install/configure it. SQL does not require it. The Quorum and T-Logs should be on RAID 1 LUNs. Separate one of course. Your databases can go on RAID 5 or 1+0 nicely. Cheers, Rod MVP - Windows Server - Clustering http://www.nw-america.com - Clustering http://www.msmvps.com/clustering - Blog [quoted text, click to view] "Randy" <Randy@discussions.microsoft.com> wrote in message news:C63B0B49-2C70-4489-B38B-94C90343660B@microsoft.com... > What's the best physical drive configuration for a SQL cluster on a SAN? > Would creating a large (14drive/2buses) physcial RAID 1+0 array with > appropriate sized LUNs be acceptable for the quorum, tempdb, ms-dtc, and > sql > tlog shared resources? > > Or would individual mirrored drives defined as LUNs be better? > > Thanks, > > Randy Geyer
Years ago Tempdb could be put into RAM. Microsoft took this feature away
cause too many customers were not getting a benefit from it, they actually had worse performance. While you can indeed move Tempdb, you have to understand if your application(s) will really benefit. If you move it to the SAN on different LUN's, are they really different SAN disks or just different parts of the same SAN disks. So, if your application(s) will benefit, will the SAN disks really be different disks? If yes to both, I say go for it. In most cases, one or more answers are no, so don't move it. Cheers, Rod MVP - Windows Server - Clustering http://www.nw-america.com - Clustering http://www.msmvps.com/clustering - Blog [quoted text, click to view] "Mike Hodgson" <mwh_junk@hotmail.com> wrote in message news:eenXBI30EHA.1740@TK2MSFTNGP15.phx.gbl... >I would think there's no harm in putting tempdb on its own disk and still >many benefits as it's often the most active DB on a SQL box especially in >terms of writes (in which case the more spindles in the LUN the better). > > "Rodney R. Fournier [MVP]" <rod@die.spam.die.nw-america.com> wrote in > message news:umcOjf10EHA.2608@TK2MSFTNGP10.phx.gbl... >> With SQL 2000 you don't separate tempdb anymore. >> >> If you use MSDTC that should be on it's own disk - if you don't use it >> for any applications - don't install/configure it. SQL does not require >> it. >> >> The Quorum and T-Logs should be on RAID 1 LUNs. Separate one of course. >> >> Your databases can go on RAID 5 or 1+0 nicely. >> >> Cheers, >> >> Rod >> >> MVP - Windows Server - Clustering >> http://www.nw-america.com - Clustering >> http://www.msmvps.com/clustering - Blog >> >> "Randy" <Randy@discussions.microsoft.com> wrote in message >> news:C63B0B49-2C70-4489-B38B-94C90343660B@microsoft.com... >>> What's the best physical drive configuration for a SQL cluster on a SAN? >>> Would creating a large (14drive/2buses) physcial RAID 1+0 array with >>> appropriate sized LUNs be acceptable for the quorum, tempdb, ms-dtc, and >>> sql >>> tlog shared resources? >>> >>> Or would individual mirrored drives defined as LUNs be better? >>> >>> Thanks, >>> >>> Randy Geyer >> >> > >
If the availability of disk space is first priority, try
creating 2 arrays: one containing RAID 5 LUNs, and the other containing RAID 0+1 LUNs. Then plan the LUN allocated for data and install on RAID 5 LUN while transaction log on 0+1 LUN. ^^ Regards, Alfred XYZ [quoted text, click to view] >-----Original Message-----
>Running RAID-10 (0+1/1+0) for all drives will be a better benefit than >RAID-5, far outweighing TempDb on a separate drive. > >Regards >-------------------------------- >Mike Epprecht, Microsoft SQL Server MVP >Zurich, Switzerland > >IM: mike@epprecht.net > >MVP Program: http://www.microsoft.com/mvp > >Blog: http://www.msmvps.com/epprecht/ > >"Rodney R. Fournier [MVP]" <rod@die.spam.die.nw- america.com> wrote in >message news:uz$vmU30EHA.3336@TK2MSFTNGP11.phx.gbl... >> Years ago Tempdb could be put into RAM. Microsoft took this feature away >> cause too many customers were not getting a benefit from it, they actually >> had worse performance. While you can indeed move Tempdb, you have to >> understand if your application(s) will really benefit. If you move it to >the >> SAN on different LUN's, are they really different SAN disks or just >> different parts of the same SAN disks. >> >> So, if your application(s) will benefit, will the SAN disks really be >> different disks? If yes to both, I say go for it. In most cases, one or >more >> answers are no, so don't move it. >> >> Cheers, >> >> Rod >> >> MVP - Windows Server - Clustering >> http://www.nw-america.com - Clustering >> http://www.msmvps.com/clustering - Blog >> >> "Mike Hodgson" <mwh_junk@hotmail.com> wrote in message >> news:eenXBI30EHA.1740@TK2MSFTNGP15.phx.gbl... >> >I would think there's no harm in putting tempdb on its own disk and still >> >many benefits as it's often the most active DB on a SQL box especially in >> >terms of writes (in which case the more spindles in the LUN the better). >> > >> > "Rodney R. Fournier [MVP]" <rod@die.spam.die.nw- america.com> wrote in >> > message news:umcOjf10EHA.2608@TK2MSFTNGP10.phx.gbl... >> >> With SQL 2000 you don't separate tempdb anymore. >> >> >> >> If you use MSDTC that should be on it's own disk - if you don't use it >> >> for any applications - don't install/configure it. SQL does not require >> >> it. >> >> >> >> The Quorum and T-Logs should be on RAID 1 LUNs. Separate one of course. >> >> >> >> Your databases can go on RAID 5 or 1+0 nicely. >> >> >> >> Cheers, >> >> >> >> Rod >> >> >> >> MVP - Windows Server - Clustering >> >> http://www.nw-america.com - Clustering >> >> http://www.msmvps.com/clustering - Blog >> >> >> >> "Randy" <Randy@discussions.microsoft.com> wrote in message >> >> news:C63B0B49-2C70-4489-B38B- 94C90343660B@microsoft.com... >> >>> What's the best physical drive configuration for a SQL cluster on a >SAN? >> >>> Would creating a large (14drive/2buses) physcial RAID 1+0 array with >> >>> appropriate sized LUNs be acceptable for the quorum, tempdb, ms-dtc, >and >> >>> sql >> >>> tlog shared resources? >> >>> >> >>> Or would individual mirrored drives defined as LUNs be better? >> >>> >> >>> Thanks, >> >>> >> >>> Randy Geyer >> >> >> >> >> > >> > >> >> > > >.
Running RAID-10 (0+1/1+0) for all drives will be a better benefit than
RAID-5, far outweighing TempDb on a separate drive. Regards -------------------------------- Mike Epprecht, Microsoft SQL Server MVP Zurich, Switzerland IM: mike@epprecht.net MVP Program: http://www.microsoft.com/mvp Blog: http://www.msmvps.com/epprecht/ "Rodney R. Fournier [MVP]" <rod@die.spam.die.nw-america.com> wrote in message news:uz$vmU30EHA.3336@TK2MSFTNGP11.phx.gbl... [quoted text, click to view] > Years ago Tempdb could be put into RAM. Microsoft took this feature away > cause too many customers were not getting a benefit from it, they actually > had worse performance. While you can indeed move Tempdb, you have to > understand if your application(s) will really benefit. If you move it to the > SAN on different LUN's, are they really different SAN disks or just > different parts of the same SAN disks. > > So, if your application(s) will benefit, will the SAN disks really be > different disks? If yes to both, I say go for it. In most cases, one or more > answers are no, so don't move it. > > Cheers, > > Rod > > MVP - Windows Server - Clustering > http://www.nw-america.com - Clustering > http://www.msmvps.com/clustering - Blog > > "Mike Hodgson" <mwh_junk@hotmail.com> wrote in message > news:eenXBI30EHA.1740@TK2MSFTNGP15.phx.gbl... > >I would think there's no harm in putting tempdb on its own disk and still > >many benefits as it's often the most active DB on a SQL box especially in > >terms of writes (in which case the more spindles in the LUN the better). > > > > "Rodney R. Fournier [MVP]" <rod@die.spam.die.nw-america.com> wrote in > > message news:umcOjf10EHA.2608@TK2MSFTNGP10.phx.gbl... > >> With SQL 2000 you don't separate tempdb anymore. > >> > >> If you use MSDTC that should be on it's own disk - if you don't use it > >> for any applications - don't install/configure it. SQL does not require > >> it. > >> > >> The Quorum and T-Logs should be on RAID 1 LUNs. Separate one of course. > >> > >> Your databases can go on RAID 5 or 1+0 nicely. > >> > >> Cheers, > >> > >> Rod > >> > >> MVP - Windows Server - Clustering > >> http://www.nw-america.com - Clustering > >> http://www.msmvps.com/clustering - Blog > >> > >> "Randy" <Randy@discussions.microsoft.com> wrote in message > >> news:C63B0B49-2C70-4489-B38B-94C90343660B@microsoft.com... > >>> What's the best physical drive configuration for a SQL cluster on a SAN? > >>> Would creating a large (14drive/2buses) physcial RAID 1+0 array with > >>> appropriate sized LUNs be acceptable for the quorum, tempdb, ms-dtc, and > >>> sql > >>> tlog shared resources? > >>> > >>> Or would individual mirrored drives defined as LUNs be better? > >>> > >>> Thanks, > >>> > >>> Randy Geyer > >> > >> > > > > > >
I would think there's no harm in putting tempdb on its own disk and still
many benefits as it's often the most active DB on a SQL box especially in terms of writes (in which case the more spindles in the LUN the better). "Rodney R. Fournier [MVP]" <rod@die.spam.die.nw-america.com> wrote in message news:umcOjf10EHA.2608@TK2MSFTNGP10.phx.gbl... [quoted text, click to view] > With SQL 2000 you don't separate tempdb anymore. > > If you use MSDTC that should be on it's own disk - if you don't use it for > any applications - don't install/configure it. SQL does not require it. > > The Quorum and T-Logs should be on RAID 1 LUNs. Separate one of course. > > Your databases can go on RAID 5 or 1+0 nicely. > > Cheers, > > Rod > > MVP - Windows Server - Clustering > http://www.nw-america.com - Clustering > http://www.msmvps.com/clustering - Blog > > "Randy" <Randy@discussions.microsoft.com> wrote in message > news:C63B0B49-2C70-4489-B38B-94C90343660B@microsoft.com... >> What's the best physical drive configuration for a SQL cluster on a SAN? >> Would creating a large (14drive/2buses) physcial RAID 1+0 array with >> appropriate sized LUNs be acceptable for the quorum, tempdb, ms-dtc, and >> sql >> tlog shared resources? >> >> Or would individual mirrored drives defined as LUNs be better? >> >> Thanks, >> >> Randy Geyer > >
Agreed.
"Rodney R. Fournier [MVP]" <rod@die.spam.die.nw-america.com> wrote in message news:uz$vmU30EHA.3336@TK2MSFTNGP11.phx.gbl... [quoted text, click to view] > Years ago Tempdb could be put into RAM. Microsoft took this feature away > cause too many customers were not getting a benefit from it, they actually > had worse performance. While you can indeed move Tempdb, you have to > understand if your application(s) will really benefit. If you move it to > the SAN on different LUN's, are they really different SAN disks or just > different parts of the same SAN disks. > > So, if your application(s) will benefit, will the SAN disks really be > different disks? If yes to both, I say go for it. In most cases, one or > more answers are no, so don't move it. > > Cheers, > > Rod > > MVP - Windows Server - Clustering > http://www.nw-america.com - Clustering > http://www.msmvps.com/clustering - Blog > > "Mike Hodgson" <mwh_junk@hotmail.com> wrote in message > news:eenXBI30EHA.1740@TK2MSFTNGP15.phx.gbl... >>I would think there's no harm in putting tempdb on its own disk and still >>many benefits as it's often the most active DB on a SQL box especially in >>terms of writes (in which case the more spindles in the LUN the better). >> >> "Rodney R. Fournier [MVP]" <rod@die.spam.die.nw-america.com> wrote in >> message news:umcOjf10EHA.2608@TK2MSFTNGP10.phx.gbl... >>> With SQL 2000 you don't separate tempdb anymore. >>> >>> If you use MSDTC that should be on it's own disk - if you don't use it >>> for any applications - don't install/configure it. SQL does not require >>> it. >>> >>> The Quorum and T-Logs should be on RAID 1 LUNs. Separate one of course. >>> >>> Your databases can go on RAID 5 or 1+0 nicely. >>> >>> Cheers, >>> >>> Rod >>> >>> MVP - Windows Server - Clustering >>> http://www.nw-america.com - Clustering >>> http://www.msmvps.com/clustering - Blog >>> >>> "Randy" <Randy@discussions.microsoft.com> wrote in message >>> news:C63B0B49-2C70-4489-B38B-94C90343660B@microsoft.com... >>>> What's the best physical drive configuration for a SQL cluster on a >>>> SAN? >>>> Would creating a large (14drive/2buses) physcial RAID 1+0 array with >>>> appropriate sized LUNs be acceptable for the quorum, tempdb, ms-dtc, >>>> and sql >>>> tlog shared resources? >>>> >>>> Or would individual mirrored drives defined as LUNs be better? >>>> >>>> Thanks, >>>> >>>> Randy Geyer >>> >>> >> >> > >
Use RAID 5 on read-only volumes. Any disk volume that does more than 10 percent writes is not a good candidate for RAID 5.
Use RAID 10 whenever the array experiences more than 10 percent writes. In general, for OLTP systems, RAID 10 is the best for data + logs. Regarding the original qs - as many have already answered --- have seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups etc. One huge lun is not recommended for high availability. Best Regards, Uttam Parui Microsoft Corporation This posting is provided "AS IS" with no warranties, and confers no rights. Are you secure? For information about the Strategic Technology Protection Program and to order your FREE Security Tool Kit, please visit http://www.microsoft.com/security. Microsoft highly recommends that users with Internet access update their Microsoft software to better protect against viruses and security vulnerabilities. The easiest way to do this is to visit the following websites: http://www.microsoft.com/protect http://www.microsoft.com/security/guidance/default.mspx
Thank you Uttam. You came very close to answering my original qs. :)
My question is: I'd like to create a large physical raid 1+0 array with 12 physical drives on the SAN. Then, partition that array into several LUNs for the shared resources (quorum, logs, ms-dtc, etc). Space is not an issue. Is that configuration more performant than individual RAID 1 arrays with 2 drives each for each shared resource? Thanks! - rg [quoted text, click to view] "Uttam Parui[MS]" wrote:
> Use RAID 5 on read-only volumes. Any disk volume that does more than 10 percent writes is not a good candidate for RAID 5. > Use RAID 10 whenever the array experiences more than 10 percent writes. > > In general, for OLTP systems, RAID 10 is the best for data + logs. > > Regarding the original qs - as many have already answered --- have seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups etc. One huge lun is not recommended for high availability. > > Best Regards, > > Uttam Parui > Microsoft Corporation > > This posting is provided "AS IS" with no warranties, and confers no rights. > > Are you secure? For information about the Strategic Technology Protection Program and to order your FREE Security Tool Kit, please visit http://www.microsoft.com/security. > > Microsoft highly recommends that users with Internet access update their Microsoft software to better protect against viruses and security vulnerabilities. The easiest way to do this is to visit the following websites: > http://www.microsoft.com/protect > http://www.microsoft.com/security/guidance/default.mspx > >
Your partitioning of the array into several LUNs must be transparent to the
OS, and the LUNs should be presented to the OS as separate SCSI (SCSI on Fibre) devices. In general, it may perform better. But it depends a lot on the ability of the SAN to handle a mixed I/O workload. For certain SAN with large and layered cache, mixing sequential log I/Os with random data I/Os may not have as big an adverse impact on the performance as it does with some other storage devices. But it is not advisable to share data files and log files on the same spindles for disaster recovery reasons. Linchi [quoted text, click to view] "Randy" <Randy@discussions.microsoft.com> wrote in message seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups etc. Onenews:ADF19E59-3ACF-43AF-917B-7308D773BE62@microsoft.com... > Thank you Uttam. You came very close to answering my original qs. :) > > My question is: I'd like to create a large physical raid 1+0 array with 12 > physical drives on the SAN. Then, partition that array into several LUNs for > the shared resources (quorum, logs, ms-dtc, etc). Space is not an issue. Is > that configuration more performant than individual RAID 1 arrays with 2 > drives each for each shared resource? > > Thanks! - rg > > "Uttam Parui[MS]" wrote: > > > Use RAID 5 on read-only volumes. Any disk volume that does more than 10 percent writes is not a good candidate for RAID 5. > > Use RAID 10 whenever the array experiences more than 10 percent writes. > > > > In general, for OLTP systems, RAID 10 is the best for data + logs. > > > > Regarding the original qs - as many have already answered --- have huge lun is not recommended for high availability. [quoted text, click to view] > > Protection Program and to order your FREE Security Tool Kit, please visit> > Best Regards, > > > > Uttam Parui > > Microsoft Corporation > > > > This posting is provided "AS IS" with no warranties, and confers no rights. > > > > Are you secure? For information about the Strategic Technology http://www.microsoft.com/security. [quoted text, click to view] > > Microsoft software to better protect against viruses and security> > Microsoft highly recommends that users with Internet access update their vulnerabilities. The easiest way to do this is to visit the following websites: [quoted text, click to view] > > http://www.microsoft.com/protect > > http://www.microsoft.com/security/guidance/default.mspx > > > > > >
I also have a question for Uttam Parui...
I plan to install 2 node cluster (initially) and then add third and may be fourth node later. I plan to do the following: 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = E, F, G 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q Each Server has following: C:\ (mirrrored) = OS D:\ (mirrored) = SQL Binary + PAGE FILE Server1 (Node1/Active) Data=E:\ Log=S:\ Quorum=Q:\ Server2 (Node2/Passive) Data=E:\ Log=S:\ Quorum:Q:\ Server3 (To be added later as an ACTIVE Node with Server 2 as PASSIVE/StandBy) Data=F:\ Log=T:\ Quorum=Q:\ First, I am not sure if this is correct for 3/4 four nodes cluster and if it is how it will work/affect FAILOVER when it happens between S1 & S2 or S3 & S2 I would greatly appreciate if you can provide any feedback. Regards, Kamal [quoted text, click to view] "Uttam Parui[MS]" wrote:
> Use RAID 5 on read-only volumes. Any disk volume that does more than 10 percent writes is not a good candidate for RAID 5. > Use RAID 10 whenever the array experiences more than 10 percent writes. > > In general, for OLTP systems, RAID 10 is the best for data + logs. > > Regarding the original qs - as many have already answered --- have seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups etc. One huge lun is not recommended for high availability. > > Best Regards, > > Uttam Parui > Microsoft Corporation > > This posting is provided "AS IS" with no warranties, and confers no rights. > > Are you secure? For information about the Strategic Technology Protection Program and to order your FREE Security Tool Kit, please visit http://www.microsoft.com/security. > > Microsoft highly recommends that users with Internet access update their Microsoft software to better protect against viruses and security vulnerabilities. The easiest way to do this is to visit the following websites: > http://www.microsoft.com/protect > http://www.microsoft.com/security/guidance/default.mspx > >
First, thanks for your response.
We plan to keep our existing backup server (lots of disk space...). I know it will go over the LAN as oppose to SAN but until we add more drives to existing SAN we plan to continue backup on backup server (online backup ) and then to TAPE. With regards to SQL Clustering do you see any issues with the proposed configuration? I think I have read some where that in multi-instance SQL clustering all nodes should have SAME drive letters. Is this true/correct? or what are the possible issues with same drive letter or the drive letter should be different across all NODES but then explain how the failover will see/work etc. Please advice.... Kamal. [quoted text, click to view] "Mike Epprecht (SQL MVP)" wrote:
> An where are you going to put your backups? > > Or are you relying on tape backup software? > > Regards > -------------------------------- > Mike Epprecht, Microsoft SQL Server MVP > Zurich, Switzerland > > IM: mike@epprecht.net > > MVP Program: http://www.microsoft.com/mvp > > Blog: http://www.msmvps.com/epprecht/ > > "Kamal Hassan" <Kamal Hassan@discussions.microsoft.com> wrote in message > news:1CF126AB-D562-4706-8E90-2A2900C56C38@microsoft.com... > > I also have a question for Uttam Parui... > > > > I plan to install 2 node cluster (initially) and then add third and may be > > fourth node later. > > > > I plan to do the following: > > > > 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = E, F, G > > 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q > > > > Each Server has following: > > > > C:\ (mirrrored) = OS > > D:\ (mirrored) = SQL Binary + PAGE FILE > > > > Server1 (Node1/Active) > > Data=E:\ > > Log=S:\ > > Quorum=Q:\ > > > > Server2 (Node2/Passive) > > Data=E:\ > > Log=S:\ > > Quorum:Q:\ > > > > Server3 (To be added later as an ACTIVE Node with Server 2 as > PASSIVE/StandBy) > > Data=F:\ > > Log=T:\ > > Quorum=Q:\ > > > > First, I am not sure if this is correct for 3/4 four nodes cluster and if > it > > is how it will work/affect FAILOVER when it happens between S1 & S2 or S3 > & S2 > > > > I would greatly appreciate if you can provide any feedback. > > > > Regards, > > > > Kamal > > > > > > > > > > "Uttam Parui[MS]" wrote: > > > > > Use RAID 5 on read-only volumes. Any disk volume that does more than 10 > percent writes is not a good candidate for RAID 5. > > > Use RAID 10 whenever the array experiences more than 10 percent writes. > > > > > > In general, for OLTP systems, RAID 10 is the best for data + logs. > > > > > > Regarding the original qs - as many have already answered --- have > seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups etc. One > huge lun is not recommended for high availability. > > > > > > Best Regards, > > > > > > Uttam Parui > > > Microsoft Corporation > > > > > > This posting is provided "AS IS" with no warranties, and confers no > rights. > > > > > > Are you secure? For information about the Strategic Technology > Protection Program and to order your FREE Security Tool Kit, please visit > http://www.microsoft.com/security. > > > > > > Microsoft highly recommends that users with Internet access update their > Microsoft software to better protect against viruses and security > vulnerabilities. The easiest way to do this is to visit the following > websites: > > > http://www.microsoft.com/protect > > > http://www.microsoft.com/security/guidance/default.mspx > > > > > > > > > > >
An where are you going to put your backups?
Or are you relying on tape backup software? Regards -------------------------------- Mike Epprecht, Microsoft SQL Server MVP Zurich, Switzerland IM: mike@epprecht.net MVP Program: http://www.microsoft.com/mvp Blog: http://www.msmvps.com/epprecht/ [quoted text, click to view] "Kamal Hassan" <Kamal Hassan@discussions.microsoft.com> wrote in message seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups etc. Onenews:1CF126AB-D562-4706-8E90-2A2900C56C38@microsoft.com... > I also have a question for Uttam Parui... > > I plan to install 2 node cluster (initially) and then add third and may be > fourth node later. > > I plan to do the following: > > 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = E, F, G > 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q > > Each Server has following: > > C:\ (mirrrored) = OS > D:\ (mirrored) = SQL Binary + PAGE FILE > > Server1 (Node1/Active) > Data=E:\ > Log=S:\ > Quorum=Q:\ > > Server2 (Node2/Passive) > Data=E:\ > Log=S:\ > Quorum:Q:\ > > Server3 (To be added later as an ACTIVE Node with Server 2 as PASSIVE/StandBy) > Data=F:\ > Log=T:\ > Quorum=Q:\ > > First, I am not sure if this is correct for 3/4 four nodes cluster and if it > is how it will work/affect FAILOVER when it happens between S1 & S2 or S3 & S2 > > I would greatly appreciate if you can provide any feedback. > > Regards, > > Kamal > > > > > "Uttam Parui[MS]" wrote: > > > Use RAID 5 on read-only volumes. Any disk volume that does more than 10 percent writes is not a good candidate for RAID 5. > > Use RAID 10 whenever the array experiences more than 10 percent writes. > > > > In general, for OLTP systems, RAID 10 is the best for data + logs. > > > > Regarding the original qs - as many have already answered --- have huge lun is not recommended for high availability. [quoted text, click to view] > > Protection Program and to order your FREE Security Tool Kit, please visit> > Best Regards, > > > > Uttam Parui > > Microsoft Corporation > > > > This posting is provided "AS IS" with no warranties, and confers no rights. > > > > Are you secure? For information about the Strategic Technology http://www.microsoft.com/security. [quoted text, click to view] > > Microsoft software to better protect against viruses and security> > Microsoft highly recommends that users with Internet access update their vulnerabilities. The easiest way to do this is to visit the following websites: [quoted text, click to view] > > http://www.microsoft.com/protect > > http://www.microsoft.com/security/guidance/default.mspx > > > > > >
First, backup over a LAN to disk is not a problem. I do it all the time.
You may want a dedicated link to avoid saturating the LAN. Second, All host nodes in a cluster see the same drives with the same drive letters. Therefore, you must plan drive letters on a cluster-wide basis. The Cluster software arbitrates who can actually mount and control the device to avoid data corruption. -- Geoff N. Hiten Microsoft SQL Server MVP Senior Database Administrator Careerbuilder.com I support the Professional Association for SQL Server www.sqlpass.org [quoted text, click to view] "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message news:3CDFE4C7-D46F-44AF-80F2-6E1204DC17FF@microsoft.com... > First, thanks for your response. > > We plan to keep our existing backup server (lots of disk space...). I know > it will go over the LAN as oppose to SAN but until we add more drives to > existing SAN we plan to continue backup on backup server (online backup ) and > then to TAPE. > > With regards to SQL Clustering do you see any issues with the proposed > configuration? I think I have read some where that in multi-instance SQL > clustering all nodes should have SAME drive letters. Is this true/correct? or > what are the possible issues with same drive letter or the drive letter > should be different across all NODES but then explain how the failover will > see/work etc. > > Please advice.... > > Kamal. > > "Mike Epprecht (SQL MVP)" wrote: > > > An where are you going to put your backups? > > > > Or are you relying on tape backup software? > > > > Regards > > -------------------------------- > > Mike Epprecht, Microsoft SQL Server MVP > > Zurich, Switzerland > > > > IM: mike@epprecht.net > > > > MVP Program: http://www.microsoft.com/mvp > > > > Blog: http://www.msmvps.com/epprecht/ > > > > "Kamal Hassan" <Kamal Hassan@discussions.microsoft.com> wrote in message > > news:1CF126AB-D562-4706-8E90-2A2900C56C38@microsoft.com... > > > I also have a question for Uttam Parui... > > > > > > I plan to install 2 node cluster (initially) and then add third and may be > > > fourth node later. > > > > > > I plan to do the following: > > > > > > 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = E, F, G > > > 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q > > > > > > Each Server has following: > > > > > > C:\ (mirrrored) = OS > > > D:\ (mirrored) = SQL Binary + PAGE FILE > > > > > > Server1 (Node1/Active) > > > Data=E:\ > > > Log=S:\ > > > Quorum=Q:\ > > > > > > Server2 (Node2/Passive) > > > Data=E:\ > > > Log=S:\ > > > Quorum:Q:\ > > > > > > Server3 (To be added later as an ACTIVE Node with Server 2 as > > PASSIVE/StandBy) > > > Data=F:\ > > > Log=T:\ > > > Quorum=Q:\ > > > > > > First, I am not sure if this is correct for 3/4 four nodes cluster and if > > it > > > is how it will work/affect FAILOVER when it happens between S1 & S2 or S3 > > & S2 > > > > > > I would greatly appreciate if you can provide any feedback. > > > > > > Regards, > > > > > > Kamal > > > > > > > > > > > > > > > "Uttam Parui[MS]" wrote: > > > > > > > Use RAID 5 on read-only volumes. Any disk volume that does more than 10 > > percent writes is not a good candidate for RAID 5. > > > > Use RAID 10 whenever the array experiences more than 10 percent writes. > > > > > > > > In general, for OLTP systems, RAID 10 is the best for data + logs. > > > > > > > > Regarding the original qs - as many have already answered --- have > > seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups etc. One > > huge lun is not recommended for high availability. > > > > > > > > Best Regards, > > > > > > > > Uttam Parui > > > > Microsoft Corporation > > > > > > > > This posting is provided "AS IS" with no warranties, and confers no > > rights. > > > > > > > > Are you secure? For information about the Strategic Technology > > Protection Program and to order your FREE Security Tool Kit, please visit > > http://www.microsoft.com/security. > > > > > > > > Microsoft highly recommends that users with Internet access update their > > Microsoft software to better protect against viruses and security > > vulnerabilities. The easiest way to do this is to visit the following > > websites: > > > > http://www.microsoft.com/protect > > > > http://www.microsoft.com/security/guidance/default.mspx > > > > > > > > > > > > > > > > > >
Hi,
First, thanks for your responses which helped me to setup/running SQL cluster in SAN. Now I have another question regarding SQL Server Clustering and SAN. We plan to have 2-instance (1 StandBy) Cluster running on SAN. What would be the best RAID configuration using SAN. Disk Drives/Space not an issue? Any comments/feedback would be greatly appreciated. Kamal. [quoted text, click to view] "Geoff N. Hiten" wrote:
> First, backup over a LAN to disk is not a problem. I do it all the time. > You may want a dedicated link to avoid saturating the LAN. > > Second, All host nodes in a cluster see the same drives with the same drive > letters. Therefore, you must plan drive letters on a cluster-wide basis. > The Cluster software arbitrates who can actually mount and control the > device to avoid data corruption. > > -- > Geoff N. Hiten > Microsoft SQL Server MVP > Senior Database Administrator > Careerbuilder.com > > I support the Professional Association for SQL Server > www.sqlpass.org > > "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message > news:3CDFE4C7-D46F-44AF-80F2-6E1204DC17FF@microsoft.com... > > First, thanks for your response. > > > > We plan to keep our existing backup server (lots of disk space...). I know > > it will go over the LAN as oppose to SAN but until we add more drives to > > existing SAN we plan to continue backup on backup server (online backup ) > and > > then to TAPE. > > > > With regards to SQL Clustering do you see any issues with the proposed > > configuration? I think I have read some where that in multi-instance SQL > > clustering all nodes should have SAME drive letters. Is this true/correct? > or > > what are the possible issues with same drive letter or the drive letter > > should be different across all NODES but then explain how the failover > will > > see/work etc. > > > > Please advice.... > > > > Kamal. > > > > "Mike Epprecht (SQL MVP)" wrote: > > > > > An where are you going to put your backups? > > > > > > Or are you relying on tape backup software? > > > > > > Regards > > > -------------------------------- > > > Mike Epprecht, Microsoft SQL Server MVP > > > Zurich, Switzerland > > > > > > IM: mike@epprecht.net > > > > > > MVP Program: http://www.microsoft.com/mvp > > > > > > Blog: http://www.msmvps.com/epprecht/ > > > > > > "Kamal Hassan" <Kamal Hassan@discussions.microsoft.com> wrote in message > > > news:1CF126AB-D562-4706-8E90-2A2900C56C38@microsoft.com... > > > > I also have a question for Uttam Parui... > > > > > > > > I plan to install 2 node cluster (initially) and then add third and > may be > > > > fourth node later. > > > > > > > > I plan to do the following: > > > > > > > > 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = E, F, > G > > > > 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q > > > > > > > > Each Server has following: > > > > > > > > C:\ (mirrrored) = OS > > > > D:\ (mirrored) = SQL Binary + PAGE FILE > > > > > > > > Server1 (Node1/Active) > > > > Data=E:\ > > > > Log=S:\ > > > > Quorum=Q:\ > > > > > > > > Server2 (Node2/Passive) > > > > Data=E:\ > > > > Log=S:\ > > > > Quorum:Q:\ > > > > > > > > Server3 (To be added later as an ACTIVE Node with Server 2 as > > > PASSIVE/StandBy) > > > > Data=F:\ > > > > Log=T:\ > > > > Quorum=Q:\ > > > > > > > > First, I am not sure if this is correct for 3/4 four nodes cluster and > if > > > it > > > > is how it will work/affect FAILOVER when it happens between S1 & S2 or > S3 > > > & S2 > > > > > > > > I would greatly appreciate if you can provide any feedback. > > > > > > > > Regards, > > > > > > > > Kamal > > > > > > > > > > > > > > > > > > > > "Uttam Parui[MS]" wrote: > > > > > > > > > Use RAID 5 on read-only volumes. Any disk volume that does more than > 10 > > > percent writes is not a good candidate for RAID 5. > > > > > Use RAID 10 whenever the array experiences more than 10 percent > writes. > > > > > > > > > > In general, for OLTP systems, RAID 10 is the best for data + logs. > > > > > > > > > > Regarding the original qs - as many have already answered --- have > > > seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups etc. > One > > > huge lun is not recommended for high availability. > > > > > > > > > > Best Regards, > > > > > > > > > > Uttam Parui > > > > > Microsoft Corporation > > > > > > > > > > This posting is provided "AS IS" with no warranties, and confers no > > > rights. > > > > > > > > > > Are you secure? For information about the Strategic Technology > > > Protection Program and to order your FREE Security Tool Kit, please > visit > > > http://www.microsoft.com/security. > > > > > > > > > > Microsoft highly recommends that users with Internet access update > their > > > Microsoft software to better protect against viruses and security > > > vulnerabilities. The easiest way to do this is to visit the following > > > websites: > > > > > http://www.microsoft.com/protect > > > > > http://www.microsoft.com/security/guidance/default.mspx > > > > > > > > > > > > > > > > > > > > > > > > > >
Hi
RAID-10 is the fastest for both reads and writes. The more spindles the better. It depends on your SAN on how many drives it can take, but if it can take 20, in effect, 10 will be available for storage whilst the other 10 are the mirrors. Regards -------------------------------- Mike Epprecht, Microsoft SQL Server MVP Zurich, Switzerland IM: mike@epprecht.net MVP Program: http://www.microsoft.com/mvp Blog: http://www.msmvps.com/epprecht/ [quoted text, click to view] "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message news:BE876DAA-1741-465D-8212-748D4CF06E1D@microsoft.com... > Hi, > > First, thanks for your responses which helped me to setup/running SQL > cluster in SAN. Now I have another question regarding SQL Server > Clustering > and SAN. > > We plan to have 2-instance (1 StandBy) Cluster running on SAN. > > What would be the best RAID configuration using SAN. > > Disk Drives/Space not an issue? > > Any comments/feedback would be greatly appreciated. > > Kamal. > > > > > > > > > > > > "Geoff N. Hiten" wrote: > >> First, backup over a LAN to disk is not a problem. I do it all the time. >> You may want a dedicated link to avoid saturating the LAN. >> >> Second, All host nodes in a cluster see the same drives with the same >> drive >> letters. Therefore, you must plan drive letters on a cluster-wide basis. >> The Cluster software arbitrates who can actually mount and control the >> device to avoid data corruption. >> >> -- >> Geoff N. Hiten >> Microsoft SQL Server MVP >> Senior Database Administrator >> Careerbuilder.com >> >> I support the Professional Association for SQL Server >> www.sqlpass.org >> >> "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message >> news:3CDFE4C7-D46F-44AF-80F2-6E1204DC17FF@microsoft.com... >> > First, thanks for your response. >> > >> > We plan to keep our existing backup server (lots of disk space...). I >> > know >> > it will go over the LAN as oppose to SAN but until we add more drives >> > to >> > existing SAN we plan to continue backup on backup server (online >> > backup ) >> and >> > then to TAPE. >> > >> > With regards to SQL Clustering do you see any issues with the proposed >> > configuration? I think I have read some where that in multi-instance >> > SQL >> > clustering all nodes should have SAME drive letters. Is this >> > true/correct? >> or >> > what are the possible issues with same drive letter or the drive letter >> > should be different across all NODES but then explain how the failover >> will >> > see/work etc. >> > >> > Please advice.... >> > >> > Kamal. >> > >> > "Mike Epprecht (SQL MVP)" wrote: >> > >> > > An where are you going to put your backups? >> > > >> > > Or are you relying on tape backup software? >> > > >> > > Regards >> > > -------------------------------- >> > > Mike Epprecht, Microsoft SQL Server MVP >> > > Zurich, Switzerland >> > > >> > > IM: mike@epprecht.net >> > > >> > > MVP Program: http://www.microsoft.com/mvp >> > > >> > > Blog: http://www.msmvps.com/epprecht/ >> > > >> > > "Kamal Hassan" <Kamal Hassan@discussions.microsoft.com> wrote in >> > > message >> > > news:1CF126AB-D562-4706-8E90-2A2900C56C38@microsoft.com... >> > > > I also have a question for Uttam Parui... >> > > > >> > > > I plan to install 2 node cluster (initially) and then add third and >> may be >> > > > fourth node later. >> > > > >> > > > I plan to do the following: >> > > > >> > > > 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = E, >> > > > F, >> G >> > > > 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q >> > > > >> > > > Each Server has following: >> > > > >> > > > C:\ (mirrrored) = OS >> > > > D:\ (mirrored) = SQL Binary + PAGE FILE >> > > > >> > > > Server1 (Node1/Active) >> > > > Data=E:\ >> > > > Log=S:\ >> > > > Quorum=Q:\ >> > > > >> > > > Server2 (Node2/Passive) >> > > > Data=E:\ >> > > > Log=S:\ >> > > > Quorum:Q:\ >> > > > >> > > > Server3 (To be added later as an ACTIVE Node with Server 2 as >> > > PASSIVE/StandBy) >> > > > Data=F:\ >> > > > Log=T:\ >> > > > Quorum=Q:\ >> > > > >> > > > First, I am not sure if this is correct for 3/4 four nodes cluster >> > > > and >> if >> > > it >> > > > is how it will work/affect FAILOVER when it happens between S1 & S2 >> > > > or >> S3 >> > > & S2 >> > > > >> > > > I would greatly appreciate if you can provide any feedback. >> > > > >> > > > Regards, >> > > > >> > > > Kamal >> > > > >> > > > >> > > > >> > > > >> > > > "Uttam Parui[MS]" wrote: >> > > > >> > > > > Use RAID 5 on read-only volumes. Any disk volume that does more >> > > > > than >> 10 >> > > percent writes is not a good candidate for RAID 5. >> > > > > Use RAID 10 whenever the array experiences more than 10 percent >> writes. >> > > > > >> > > > > In general, for OLTP systems, RAID 10 is the best for data + >> > > > > logs. >> > > > > >> > > > > Regarding the original qs - as many have already answered --- >> > > > > have >> > > seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups >> > > etc. >> One >> > > huge lun is not recommended for high availability. >> > > > > >> > > > > Best Regards, >> > > > > >> > > > > Uttam Parui >> > > > > Microsoft Corporation >> > > > > >> > > > > This posting is provided "AS IS" with no warranties, and confers >> > > > > no >> > > rights. >> > > > > >> > > > > Are you secure? For information about the Strategic Technology >> > > Protection Program and to order your FREE Security Tool Kit, please >> visit >> > > http://www.microsoft.com/security. >> > > > > >> > > > > Microsoft highly recommends that users with Internet access >> > > > > update >> their >> > > Microsoft software to better protect against viruses and security >> > > vulnerabilities. The easiest way to do this is to visit the following >> > > websites: >> > > > > http://www.microsoft.com/protect >> > > > > http://www.microsoft.com/security/guidance/default.mspx >> > > > > >> > > > > >> > > > > >> > > >> > > >> > > >> >> >>
Thanks Mike,
We are using (Lab/Test) EMC-CX300. The first enclosure has a capacity of 15 divres. Currently the configuration looks like this: Windows 2003 Enterprise Edition SQL 2000 Enterprise Edition SQL Clustering (currenlty configured for 1 instance, plan to add two more) a) 6 drives (three RAID-5 LUNS, 100 GB each for Data for three INSTANCES) b) 4 drives (three RAID-1 LUNS, 33 GB, 28 B (Log) 5.0 GB for QUORUM) We are discussing/debating weather should we go with RAID 10 or not. As we understand and you also stated the RAID 10 is the fastest we plan to go that route but not sure how to CARVE the disk/luns for BEST performance for 3 SQL Server instances. I would really appreciate your comments/feedback. Thanks. Kamal. [quoted text, click to view] "Mike Epprecht (SQL MVP)" wrote:
> Hi > > RAID-10 is the fastest for both reads and writes. The more spindles the > better. > > It depends on your SAN on how many drives it can take, but if it can take > 20, in effect, 10 will be available for storage whilst the other 10 are the > mirrors. > > Regards > -------------------------------- > Mike Epprecht, Microsoft SQL Server MVP > Zurich, Switzerland > > IM: mike@epprecht.net > > MVP Program: http://www.microsoft.com/mvp > > Blog: http://www.msmvps.com/epprecht/ > > "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message > news:BE876DAA-1741-465D-8212-748D4CF06E1D@microsoft.com... > > Hi, > > > > First, thanks for your responses which helped me to setup/running SQL > > cluster in SAN. Now I have another question regarding SQL Server > > Clustering > > and SAN. > > > > We plan to have 2-instance (1 StandBy) Cluster running on SAN. > > > > What would be the best RAID configuration using SAN. > > > > Disk Drives/Space not an issue? > > > > Any comments/feedback would be greatly appreciated. > > > > Kamal. > > > > > > > > > > > > > > > > > > > > > > > > "Geoff N. Hiten" wrote: > > > >> First, backup over a LAN to disk is not a problem. I do it all the time. > >> You may want a dedicated link to avoid saturating the LAN. > >> > >> Second, All host nodes in a cluster see the same drives with the same > >> drive > >> letters. Therefore, you must plan drive letters on a cluster-wide basis. > >> The Cluster software arbitrates who can actually mount and control the > >> device to avoid data corruption. > >> > >> -- > >> Geoff N. Hiten > >> Microsoft SQL Server MVP > >> Senior Database Administrator > >> Careerbuilder.com > >> > >> I support the Professional Association for SQL Server > >> www.sqlpass.org > >> > >> "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message > >> news:3CDFE4C7-D46F-44AF-80F2-6E1204DC17FF@microsoft.com... > >> > First, thanks for your response. > >> > > >> > We plan to keep our existing backup server (lots of disk space...). I > >> > know > >> > it will go over the LAN as oppose to SAN but until we add more drives > >> > to > >> > existing SAN we plan to continue backup on backup server (online > >> > backup ) > >> and > >> > then to TAPE. > >> > > >> > With regards to SQL Clustering do you see any issues with the proposed > >> > configuration? I think I have read some where that in multi-instance > >> > SQL > >> > clustering all nodes should have SAME drive letters. Is this > >> > true/correct? > >> or > >> > what are the possible issues with same drive letter or the drive letter > >> > should be different across all NODES but then explain how the failover > >> will > >> > see/work etc. > >> > > >> > Please advice.... > >> > > >> > Kamal. > >> > > >> > "Mike Epprecht (SQL MVP)" wrote: > >> > > >> > > An where are you going to put your backups? > >> > > > >> > > Or are you relying on tape backup software? > >> > > > >> > > Regards > >> > > -------------------------------- > >> > > Mike Epprecht, Microsoft SQL Server MVP > >> > > Zurich, Switzerland > >> > > > >> > > IM: mike@epprecht.net > >> > > > >> > > MVP Program: http://www.microsoft.com/mvp > >> > > > >> > > Blog: http://www.msmvps.com/epprecht/ > >> > > > >> > > "Kamal Hassan" <Kamal Hassan@discussions.microsoft.com> wrote in > >> > > message > >> > > news:1CF126AB-D562-4706-8E90-2A2900C56C38@microsoft.com... > >> > > > I also have a question for Uttam Parui... > >> > > > > >> > > > I plan to install 2 node cluster (initially) and then add third and > >> may be > >> > > > fourth node later. > >> > > > > >> > > > I plan to do the following: > >> > > > > >> > > > 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = E, > >> > > > F, > >> G > >> > > > 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q > >> > > > > >> > > > Each Server has following: > >> > > > > >> > > > C:\ (mirrrored) = OS > >> > > > D:\ (mirrored) = SQL Binary + PAGE FILE > >> > > > > >> > > > Server1 (Node1/Active) > >> > > > Data=E:\ > >> > > > Log=S:\ > >> > > > Quorum=Q:\ > >> > > > > >> > > > Server2 (Node2/Passive) > >> > > > Data=E:\ > >> > > > Log=S:\ > >> > > > Quorum:Q:\ > >> > > > > >> > > > Server3 (To be added later as an ACTIVE Node with Server 2 as > >> > > PASSIVE/StandBy) > >> > > > Data=F:\ > >> > > > Log=T:\ > >> > > > Quorum=Q:\ > >> > > > > >> > > > First, I am not sure if this is correct for 3/4 four nodes cluster > >> > > > and > >> if > >> > > it > >> > > > is how it will work/affect FAILOVER when it happens between S1 & S2 > >> > > > or > >> S3 > >> > > & S2 > >> > > > > >> > > > I would greatly appreciate if you can provide any feedback. > >> > > > > >> > > > Regards, > >> > > > > >> > > > Kamal > >> > > > > >> > > > > >> > > > > >> > > > > >> > > > "Uttam Parui[MS]" wrote: > >> > > > > >> > > > > Use RAID 5 on read-only volumes. Any disk volume that does more > >> > > > > than > >> 10 > >> > > percent writes is not a good candidate for RAID 5. > >> > > > > Use RAID 10 whenever the array experiences more than 10 percent > >> writes. > >> > > > > > >> > > > > In general, for OLTP systems, RAID 10 is the best for data + > >> > > > > logs. > >> > > > > > >> > > > > Regarding the original qs - as many have already answered --- > >> > > > > have > >> > > seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups > >> > > etc. > >> One > >> > > huge lun is not recommended for high availability. > >> > > > > > >> > > > > Best Regards, > >> > > > > > >> > > > > Uttam Parui > >> > > > > Microsoft Corporation > >> > > > > > >> > > > > This posting is provided "AS IS" with no warranties, and confers > >> > > > > no > >> > > rights. > >> > > > > > >> > > > > Are you secure? For information about the Strategic Technology > >> > > Protection Program and to order your FREE Security Tool Kit, please > >> visit > >> > > http://www.microsoft.com/security. > >> > > > >
Now that I have 2-Node SQL Cluster runing using EMC-CX300.
I have two questions: a) Currently there are two DRIVES (data & log) available for my cluster group, I aded a THIRD drive to same cluster group (addditioanl databases etc. testing...) but the new DRIVES does not appear Enterprise Manager. any idea? I have tried bringing the group off-line/online but that did not help.... b) I have additional drives (data & log) available/configured for the THIRD node. Any idea/advice if there are any known issues when adding a THIRD node to an existing cluster where existing cluster NODE1 is active NODE is Standby and when we add the third node NODE3 the NODE2 also will be Standyby for NODe3....Any advice? [quoted text, click to view] "Kamal Hassan" wrote:
> Hi, > > First, thanks for your responses which helped me to setup/running SQL > cluster in SAN. Now I have another question regarding SQL Server Clustering > and SAN. > > We plan to have 2-instance (1 StandBy) Cluster running on SAN. > > What would be the best RAID configuration using SAN. > > Disk Drives/Space not an issue? > > Any comments/feedback would be greatly appreciated. > > Kamal. > > > > > > > > > > > > "Geoff N. Hiten" wrote: > > > First, backup over a LAN to disk is not a problem. I do it all the time. > > You may want a dedicated link to avoid saturating the LAN. > > > > Second, All host nodes in a cluster see the same drives with the same drive > > letters. Therefore, you must plan drive letters on a cluster-wide basis. > > The Cluster software arbitrates who can actually mount and control the > > device to avoid data corruption. > > > > -- > > Geoff N. Hiten > > Microsoft SQL Server MVP > > Senior Database Administrator > > Careerbuilder.com > > > > I support the Professional Association for SQL Server > > www.sqlpass.org > > > > "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message > > news:3CDFE4C7-D46F-44AF-80F2-6E1204DC17FF@microsoft.com... > > > First, thanks for your response. > > > > > > We plan to keep our existing backup server (lots of disk space...). I know > > > it will go over the LAN as oppose to SAN but until we add more drives to > > > existing SAN we plan to continue backup on backup server (online backup ) > > and > > > then to TAPE. > > > > > > With regards to SQL Clustering do you see any issues with the proposed > > > configuration? I think I have read some where that in multi-instance SQL > > > clustering all nodes should have SAME drive letters. Is this true/correct? > > or > > > what are the possible issues with same drive letter or the drive letter > > > should be different across all NODES but then explain how the failover > > will > > > see/work etc. > > > > > > Please advice.... > > > > > > Kamal. > > > > > > "Mike Epprecht (SQL MVP)" wrote: > > > > > > > An where are you going to put your backups? > > > > > > > > Or are you relying on tape backup software? > > > > > > > > Regards > > > > -------------------------------- > > > > Mike Epprecht, Microsoft SQL Server MVP > > > > Zurich, Switzerland > > > > > > > > IM: mike@epprecht.net > > > > > > > > MVP Program: http://www.microsoft.com/mvp > > > > > > > > Blog: http://www.msmvps.com/epprecht/ > > > > > > > > "Kamal Hassan" <Kamal Hassan@discussions.microsoft.com> wrote in message > > > > news:1CF126AB-D562-4706-8E90-2A2900C56C38@microsoft.com... > > > > > I also have a question for Uttam Parui... > > > > > > > > > > I plan to install 2 node cluster (initially) and then add third and > > may be > > > > > fourth node later. > > > > > > > > > > I plan to do the following: > > > > > > > > > > 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = E, F, > > G > > > > > 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q > > > > > > > > > > Each Server has following: > > > > > > > > > > C:\ (mirrrored) = OS > > > > > D:\ (mirrored) = SQL Binary + PAGE FILE > > > > > > > > > > Server1 (Node1/Active) > > > > > Data=E:\ > > > > > Log=S:\ > > > > > Quorum=Q:\ > > > > > > > > > > Server2 (Node2/Passive) > > > > > Data=E:\ > > > > > Log=S:\ > > > > > Quorum:Q:\ > > > > > > > > > > Server3 (To be added later as an ACTIVE Node with Server 2 as > > > > PASSIVE/StandBy) > > > > > Data=F:\ > > > > > Log=T:\ > > > > > Quorum=Q:\ > > > > > > > > > > First, I am not sure if this is correct for 3/4 four nodes cluster and > > if > > > > it > > > > > is how it will work/affect FAILOVER when it happens between S1 & S2 or > > S3 > > > > & S2 > > > > > > > > > > I would greatly appreciate if you can provide any feedback. > > > > > > > > > > Regards, > > > > > > > > > > Kamal > > > > > > > > > > > > > > > > > > > > > > > > > "Uttam Parui[MS]" wrote: > > > > > > > > > > > Use RAID 5 on read-only volumes. Any disk volume that does more than > > 10 > > > > percent writes is not a good candidate for RAID 5. > > > > > > Use RAID 10 whenever the array experiences more than 10 percent > > writes. > > > > > > > > > > > > In general, for OLTP systems, RAID 10 is the best for data + logs. > > > > > > > > > > > > Regarding the original qs - as many have already answered --- have > > > > seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups etc. > > One > > > > huge lun is not recommended for high availability. > > > > > > > > > > > > Best Regards, > > > > > > > > > > > > Uttam Parui > > > > > > Microsoft Corporation > > > > > > > > > > > > This posting is provided "AS IS" with no warranties, and confers no > > > > rights. > > > > > > > > > > > > Are you secure? For information about the Strategic Technology > > > > Protection Program and to order your FREE Security Tool Kit, please > > visit > > > > http://www.microsoft.com/security. > > > > > > > > > > > > Microsoft highly recommends that users with Internet access update > > their > > > > Microsoft software to better protect against viruses and security > > > > vulnerabilities. The easiest way to do this is to visit the following > > > > websites: > > > > > > http://www.microsoft.com/protect > > > > > > http://www.microsoft.com/security/guidance/default.mspx > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > >
a) Take SQL Server offline, make SQL Server dependent on the new drive and
SQL Server online again. b) No issues. It is well documented in BOL on how to add a node to an existing cluster and then read the SP readme on how to apply SP's to it. Regards -------------------------------- Mike Epprecht, Microsoft SQL Server MVP Zurich, Switzerland IM: mike@epprecht.net MVP Program: http://www.microsoft.com/mvp Blog: http://www.msmvps.com/epprecht/ [quoted text, click to view] "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message
news:A9E4B1BE-A79A-4B55-910A-CB77C2A94255@microsoft.com... > Now that I have 2-Node SQL Cluster runing using EMC-CX300. > > I have two questions: > > a) Currently there are two DRIVES (data & log) available for my cluster > group, I aded a THIRD drive to same cluster group (addditioanl databases > etc. > testing...) but the new DRIVES does not appear Enterprise Manager. any > idea? > I have tried bringing the group off-line/online but that did not help.... > > b) I have additional drives (data & log) available/configured for the > THIRD > node. Any idea/advice if there are any known issues when adding a THIRD > node > to an existing cluster where existing cluster NODE1 is active NODE is > Standby > and when we add the third node NODE3 the NODE2 also will be Standyby for > NODe3....Any advice? > > > > > > > > "Kamal Hassan" wrote: > >> Hi, >> >> First, thanks for your responses which helped me to setup/running SQL >> cluster in SAN. Now I have another question regarding SQL Server >> Clustering >> and SAN. >> >> We plan to have 2-instance (1 StandBy) Cluster running on SAN. >> >> What would be the best RAID configuration using SAN. >> >> Disk Drives/Space not an issue? >> >> Any comments/feedback would be greatly appreciated. >> >> Kamal. >> >> >> >> >> >> >> >> >> >> >> >> "Geoff N. Hiten" wrote: >> >> > First, backup over a LAN to disk is not a problem. I do it all the >> > time. >> > You may want a dedicated link to avoid saturating the LAN. >> > >> > Second, All host nodes in a cluster see the same drives with the same >> > drive >> > letters. Therefore, you must plan drive letters on a cluster-wide >> > basis. >> > The Cluster software arbitrates who can actually mount and control the >> > device to avoid data corruption. >> > >> > -- >> > Geoff N. Hiten >> > Microsoft SQL Server MVP >> > Senior Database Administrator >> > Careerbuilder.com >> > >> > I support the Professional Association for SQL Server >> > www.sqlpass.org >> > >> > "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message >> > news:3CDFE4C7-D46F-44AF-80F2-6E1204DC17FF@microsoft.com... >> > > First, thanks for your response. >> > > >> > > We plan to keep our existing backup server (lots of disk space...). I >> > > know >> > > it will go over the LAN as oppose to SAN but until we add more drives >> > > to >> > > existing SAN we plan to continue backup on backup server (online >> > > backup ) >> > and >> > > then to TAPE. >> > > >> > > With regards to SQL Clustering do you see any issues with the >> > > proposed >> > > configuration? I think I have read some where that in multi-instance >> > > SQL >> > > clustering all nodes should have SAME drive letters. Is this >> > > true/correct? >> > or >> > > what are the possible issues with same drive letter or the drive >> > > letter >> > > should be different across all NODES but then explain how the >> > > failover >> > will >> > > see/work etc. >> > > >> > > Please advice.... >> > > >> > > Kamal. >> > > >> > > "Mike Epprecht (SQL MVP)" wrote: >> > > >> > > > An where are you going to put your backups? >> > > > >> > > > Or are you relying on tape backup software? >> > > > >> > > > Regards >> > > > -------------------------------- >> > > > Mike Epprecht, Microsoft SQL Server MVP >> > > > Zurich, Switzerland >> > > > >> > > > IM: mike@epprecht.net >> > > > >> > > > MVP Program: http://www.microsoft.com/mvp >> > > > >> > > > Blog: http://www.msmvps.com/epprecht/ >> > > > >> > > > "Kamal Hassan" <Kamal Hassan@discussions.microsoft.com> wrote in >> > > > message >> > > > news:1CF126AB-D562-4706-8E90-2A2900C56C38@microsoft.com... >> > > > > I also have a question for Uttam Parui... >> > > > > >> > > > > I plan to install 2 node cluster (initially) and then add third >> > > > > and >> > may be >> > > > > fourth node later. >> > > > > >> > > > > I plan to do the following: >> > > > > >> > > > > 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = >> > > > > E, F, >> > G >> > > > > 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q >> > > > > >> > > > > Each Server has following: >> > > > > >> > > > > C:\ (mirrrored) = OS >> > > > > D:\ (mirrored) = SQL Binary + PAGE FILE >> > > > > >> > > > > Server1 (Node1/Active) >> > > > > Data=E:\ >> > > > > Log=S:\ >> > > > > Quorum=Q:\ >> > > > > >> > > > > Server2 (Node2/Passive) >> > > > > Data=E:\ >> > > > > Log=S:\ >> > > > > Quorum:Q:\ >> > > > > >> > > > > Server3 (To be added later as an ACTIVE Node with Server 2 as >> > > > PASSIVE/StandBy) >> > > > > Data=F:\ >> > > > > Log=T:\ >> > > > > Quorum=Q:\ >> > > > > >> > > > > First, I am not sure if this is correct for 3/4 four nodes >> > > > > cluster and >> > if >> > > > it >> > > > > is how it will work/affect FAILOVER when it happens between S1 & >> > > > > S2 or >> > S3 >> > > > & S2 >> > > > > >> > > > > I would greatly appreciate if you can provide any feedback. >> > > > > >> > > > > Regards, >> > > > > >> > > > > Kamal >> > > > > >> > > > > >> > > > > >> > > > > >> > > > > "Uttam Parui[MS]" wrote: >> > > > > >> > > > > > Use RAID 5 on read-only volumes. Any disk volume that does more >> > > > > > than >> > 10 >> > > > percent writes is not a good candidate for RAID 5. >> > > > > > Use RAID 10 whenever the array experiences more than 10 percent >> > writes. >> > > > > > >> > > > > > In general, for OLTP systems, RAID 10 is the best for data + >> > > > > > logs. >> > > > > > >> > > > > > Regarding the original qs - as many have already answered --- >> > > > > > have >> > > > seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups >> > > > etc. >> > One >> > > > huge lun is not recommended for high availability. >> > > > > > >> > > > > > Best Regards, >> > > > > > >> > > > > > Uttam Parui >> > > > > > Microsoft Corporation >> > > > > > >> > > > > > This posting is provided "AS IS" with no warranties, and >> > > > > > confers no >> > > > rights. >> > > > > > >> > > > > > Are you secure? For information about the Strategic Technology >> > > > Protection Program and to order your FREE Security Tool Kit, please >> > visit >> > > > http://www.microsoft.com/security. >> > > > > >
Thanks.
I have another question about DB backup. We have three node SQL Cluster (Virtual Instance 1 & Virtual Instance 2, 3rd Standyby) running on SAN as follows: I also have a dedictaed LUN (Drive X:\) exposed to all three HOSTS on SAN for SQL database backup. I am able to do assign that Drive (X:\) to one of the Virtual Instance (VS1) and do backup but I am UNABLE to do backup from VS2 on same Drive (X:\) as it is owned by VS1. How do I share that Drive (X:\) among two Virtual Instances (VS1 & VS2) so I can do backup on same Drive (X:\) Please advice? Thanks. [quoted text, click to view] "Kamal Hassan" wrote:
> Hi, > > First, thanks for your responses which helped me to setup/running SQL > cluster in SAN. Now I have another question regarding SQL Server Clustering > and SAN. > > We plan to have 2-instance (1 StandBy) Cluster running on SAN. > > What would be the best RAID configuration using SAN. > > Disk Drives/Space not an issue? > > Any comments/feedback would be greatly appreciated. > > Kamal. > > > > > > > > > > > > "Geoff N. Hiten" wrote: > > > First, backup over a LAN to disk is not a problem. I do it all the time. > > You may want a dedicated link to avoid saturating the LAN. > > > > Second, All host nodes in a cluster see the same drives with the same drive > > letters. Therefore, you must plan drive letters on a cluster-wide basis. > > The Cluster software arbitrates who can actually mount and control the > > device to avoid data corruption. > > > > -- > > Geoff N. Hiten > > Microsoft SQL Server MVP > > Senior Database Administrator > > Careerbuilder.com > > > > I support the Professional Association for SQL Server > > www.sqlpass.org > > > > "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message > > news:3CDFE4C7-D46F-44AF-80F2-6E1204DC17FF@microsoft.com... > > > First, thanks for your response. > > > > > > We plan to keep our existing backup server (lots of disk space...). I know > > > it will go over the LAN as oppose to SAN but until we add more drives to > > > existing SAN we plan to continue backup on backup server (online backup ) > > and > > > then to TAPE. > > > > > > With regards to SQL Clustering do you see any issues with the proposed > > > configuration? I think I have read some where that in multi-instance SQL > > > clustering all nodes should have SAME drive letters. Is this true/correct? > > or > > > what are the possible issues with same drive letter or the drive letter > > > should be different across all NODES but then explain how the failover > > will > > > see/work etc. > > > > > > Please advice.... > > > > > > Kamal. > > > > > > "Mike Epprecht (SQL MVP)" wrote: > > > > > > > An where are you going to put your backups? > > > > > > > > Or are you relying on tape backup software? > > > > > > > > Regards > > > > -------------------------------- > > > > Mike Epprecht, Microsoft SQL Server MVP > > > > Zurich, Switzerland > > > > > > > > IM: mike@epprecht.net > > > > > > > > MVP Program: http://www.microsoft.com/mvp > > > > > > > > Blog: http://www.msmvps.com/epprecht/ > > > > > > > > "Kamal Hassan" <Kamal Hassan@discussions.microsoft.com> wrote in message > > > > news:1CF126AB-D562-4706-8E90-2A2900C56C38@microsoft.com... > > > > > I also have a question for Uttam Parui... > > > > > > > > > > I plan to install 2 node cluster (initially) and then add third and > > may be > > > > > fourth node later. > > > > > > > > > > I plan to do the following: > > > > > > > > > > 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = E, F, > > G > > > > > 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q > > > > > > > > > > Each Server has following: > > > > > > > > > > C:\ (mirrrored) = OS > > > > > D:\ (mirrored) = SQL Binary + PAGE FILE > > > > > > > > > > Server1 (Node1/Active) > > > > > Data=E:\ > > > > > Log=S:\ > > > > > Quorum=Q:\ > > > > > > > > > > Server2 (Node2/Passive) > > > > > Data=E:\ > > > > > Log=S:\ > > > > > Quorum:Q:\ > > > > > > > > > > Server3 (To be added later as an ACTIVE Node with Server 2 as > > > > PASSIVE/StandBy) > > > > > Data=F:\ > > > > > Log=T:\ > > > > > Quorum=Q:\ > > > > > > > > > > First, I am not sure if this is correct for 3/4 four nodes cluster and > > if > > > > it > > > > > is how it will work/affect FAILOVER when it happens between S1 & S2 or > > S3 > > > > & S2 > > > > > > > > > > I would greatly appreciate if you can provide any feedback. > > > > > > > > > > Regards, > > > > > > > > > > Kamal > > > > > > > > > > > > > > > > > > > > > > > > > "Uttam Parui[MS]" wrote: > > > > > > > > > > > Use RAID 5 on read-only volumes. Any disk volume that does more than > > 10 > > > > percent writes is not a good candidate for RAID 5. > > > > > > Use RAID 10 whenever the array experiences more than 10 percent > > writes. > > > > > > > > > > > > In general, for OLTP systems, RAID 10 is the best for data + logs. > > > > > > > > > > > > Regarding the original qs - as many have already answered --- have > > > > seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups etc. > > One > > > > huge lun is not recommended for high availability. > > > > > > > > > > > > Best Regards, > > > > > > > > > > > > Uttam Parui > > > > > > Microsoft Corporation > > > > > > > > > > > > This posting is provided "AS IS" with no warranties, and confers no > > > > rights. > > > > > > > > > > > > Are you secure? For information about the Strategic Technology > > > > Protection Program and to order your FREE Security Tool Kit, please > > visit > > > > http://www.microsoft.com/security. > > > > > > > > > > > > Microsoft highly recommends that users with Internet access update > > their > > > > Microsoft software to better protect against viruses and security > > > > vulnerabilities. The easiest way to do this is to visit the following > > > > websites: > > > > > > http://www.microsoft.com/protect > > > > > > http://www.microsoft.com/security/guidance/default.mspx > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > >
I typically back up to a UNC share on another machine. It is typically lots
cheaper than SAN drives and you can use an ordinary tape system to provide longer term backups. http://support.microsoft.com/default.aspx?scid=kb;en-us;555128 -- Geoff N. Hiten Senior Database Administrator Microsoft SQL Server MVP [quoted text, click to view] "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message
news:FF104574-70E9-4993-8D9D-FD866A8F3C54@microsoft.com... > Thanks. > > I have another question about DB backup. > > We have three node SQL Cluster (Virtual Instance 1 & Virtual Instance 2, > 3rd > Standyby) running on SAN as follows: > > I also have a dedictaed LUN (Drive X:\) exposed to all three HOSTS on SAN > for SQL database backup. > > I am able to do assign that Drive (X:\) to one of the Virtual Instance > (VS1) > and do backup but I am UNABLE to do backup from VS2 on same Drive (X:\) as > it > is owned by VS1. > > How do I share that Drive (X:\) among two Virtual Instances (VS1 & VS2) so > I > can do backup on same Drive (X:\) > > Please advice? > > Thanks. > > > > > > "Kamal Hassan" wrote: > >> Hi, >> >> First, thanks for your responses which helped me to setup/running SQL >> cluster in SAN. Now I have another question regarding SQL Server >> Clustering >> and SAN. >> >> We plan to have 2-instance (1 StandBy) Cluster running on SAN. >> >> What would be the best RAID configuration using SAN. >> >> Disk Drives/Space not an issue? >> >> Any comments/feedback would be greatly appreciated. >> >> Kamal. >> >> >> >> >> >> >> >> >> >> >> >> "Geoff N. Hiten" wrote: >> >> > First, backup over a LAN to disk is not a problem. I do it all the >> > time. >> > You may want a dedicated link to avoid saturating the LAN. >> > >> > Second, All host nodes in a cluster see the same drives with the same >> > drive >> > letters. Therefore, you must plan drive letters on a cluster-wide >> > basis. >> > The Cluster software arbitrates who can actually mount and control the >> > device to avoid data corruption. >> > >> > -- >> > Geoff N. Hiten >> > Microsoft SQL Server MVP >> > Senior Database Administrator >> > Careerbuilder.com >> > >> > I support the Professional Association for SQL Server >> > www.sqlpass.org >> > >> > "Kamal Hassan" <KamalHassan@discussions.microsoft.com> wrote in message >> > news:3CDFE4C7-D46F-44AF-80F2-6E1204DC17FF@microsoft.com... >> > > First, thanks for your response. >> > > >> > > We plan to keep our existing backup server (lots of disk space...). I >> > > know >> > > it will go over the LAN as oppose to SAN but until we add more drives >> > > to >> > > existing SAN we plan to continue backup on backup server (online >> > > backup ) >> > and >> > > then to TAPE. >> > > >> > > With regards to SQL Clustering do you see any issues with the >> > > proposed >> > > configuration? I think I have read some where that in multi-instance >> > > SQL >> > > clustering all nodes should have SAME drive letters. Is this >> > > true/correct? >> > or >> > > what are the possible issues with same drive letter or the drive >> > > letter >> > > should be different across all NODES but then explain how the >> > > failover >> > will >> > > see/work etc. >> > > >> > > Please advice.... >> > > >> > > Kamal. >> > > >> > > "Mike Epprecht (SQL MVP)" wrote: >> > > >> > > > An where are you going to put your backups? >> > > > >> > > > Or are you relying on tape backup software? >> > > > >> > > > Regards >> > > > -------------------------------- >> > > > Mike Epprecht, Microsoft SQL Server MVP >> > > > Zurich, Switzerland >> > > > >> > > > IM: mike@epprecht.net >> > > > >> > > > MVP Program: http://www.microsoft.com/mvp >> > > > >> > > > Blog: http://www.msmvps.com/epprecht/ >> > > > >> > > > "Kamal Hassan" <Kamal Hassan@discussions.microsoft.com> wrote in >> > > > message >> > > > news:1CF126AB-D562-4706-8E90-2A2900C56C38@microsoft.com... >> > > > > I also have a question for Uttam Parui... >> > > > > >> > > > > I plan to install 2 node cluster (initially) and then add third >> > > > > and >> > may be >> > > > > fourth node later. >> > > > > >> > > > > I plan to do the following: >> > > > > >> > > > > 3 luns (SQL Data for each NODE1 i.e Server1, Server2, Server3) = >> > > > > E, F, >> > G >> > > > > 3 luns (2 SQL log shared by S1, S2, S3, 1 for Quorum) = S, T, Q >> > > > > >> > > > > Each Server has following: >> > > > > >> > > > > C:\ (mirrrored) = OS >> > > > > D:\ (mirrored) = SQL Binary + PAGE FILE >> > > > > >> > > > > Server1 (Node1/Active) >> > > > > Data=E:\ >> > > > > Log=S:\ >> > > > > Quorum=Q:\ >> > > > > >> > > > > Server2 (Node2/Passive) >> > > > > Data=E:\ >> > > > > Log=S:\ >> > > > > Quorum:Q:\ >> > > > > >> > > > > Server3 (To be added later as an ACTIVE Node with Server 2 as >> > > > PASSIVE/StandBy) >> > > > > Data=F:\ >> > > > > Log=T:\ >> > > > > Quorum=Q:\ >> > > > > >> > > > > First, I am not sure if this is correct for 3/4 four nodes >> > > > > cluster and >> > if >> > > > it >> > > > > is how it will work/affect FAILOVER when it happens between S1 & >> > > > > S2 or >> > S3 >> > > > & S2 >> > > > > >> > > > > I would greatly appreciate if you can provide any feedback. >> > > > > >> > > > > Regards, >> > > > > >> > > > > Kamal >> > > > > >> > > > > >> > > > > >> > > > > >> > > > > "Uttam Parui[MS]" wrote: >> > > > > >> > > > > > Use RAID 5 on read-only volumes. Any disk volume that does more >> > > > > > than >> > 10 >> > > > percent writes is not a good candidate for RAID 5. >> > > > > > Use RAID 10 whenever the array experiences more than 10 percent >> > writes. >> > > > > > >> > > > > > In general, for OLTP systems, RAID 10 is the best for data + >> > > > > > logs. >> > > > > > >> > > > > > Regarding the original qs - as many have already answered --- >> > > > > > have >> > > > seperate luns for Quorum, msdtc, sql data, sql Tlogs, SQL backups >> > > > etc. >> > One >> > > > huge lun is not recommended for high availability. >> > > > > > >> > > > > > Best Regards, >> > > > > > >> > > > > > Uttam Parui >> > > > > > Microsoft Corporation >> > > > > > >> > > > > > This posting is provided "AS IS" with no warranties, and >> > > > > > confers no >> > > > rights. >> > > > > > >> > > > > > Are you secure? For information about the Strategic Technology >> > > > Protection Program and to order your FREE Security Tool Kit, please >> > visit >> > > > http://www.microsoft.com/security. >> > > > > > >> > > > > > Microsoft highly recommends that users with Internet access >> > > > > > update >> > their >> > > > Microsoft software to better protect against viruses and security >> > > > vulnerabilities. The easiest way to do this is to visit the >> > > > following >> > > > websites:
Don't see what you're looking for? Try a search.
|