mtmmza

updated

14 years ago

M

mtmmza

6 Posts

0

1381

March 5th, 2013 04:00

Problems with RM 5.4.1 and SQL 2008 Enterprise

Hello all, i try to do a consistent Snapshot of a SQL database, and i get one error.

This works a long time ago, then this was not use anymore, and now we can reuse it and is not working....

Any suggest?

Thanks!

-----

04/03/2013 16:58:44 MZ71 Replica 04/03/2013 16:58:44 created from application set SQL_Bantotal, job SQL_Consistente by Administrator.

04/03/2013 16:58:44 MZ71 Starting simulation of [application set: SQL_Bantotal / job: SQL_Consistente] at time 04/03/2013 16:58:44.

04/03/2013 16:58:44 MZ71 This operation can take a long time. Please be patient.

04/03/2013 16:59:32 MZ71 001564 INFORMATION: Need storage for 9 source device residing on VNX block CKM00114700023.

04/03/2013 16:59:32 MZ71 001566 INFORMATION: Storage Policy will consider 21 device(s) from the entire Replication Manager database.

04/03/2013 16:59:32 MZ71 001568 INFORMATION: After removing 0 in use devices, 0 devices that are not visible on the mount host and 19 devices that are the wrong storage type, there are now 2 devices being considered for this replica.

04/03/2013 16:59:33 MZ71 Original device clariion_CKM00114700023_00011 is paired with replication device clariion_CKM00114700023_MZ71-0011-1775.

04/03/2013 16:59:33 MZ71 Original device clariion_CKM00114700023_00012 is paired with replication device clariion_CKM00114700023_MZ71-0012-1775.

04/03/2013 16:59:33 MZ71 Original device clariion_CKM00114700023_00013 is paired with replication device clariion_CKM00114700023_MZ71-0013-1775.

04/03/2013 16:59:33 MZ71 Original device clariion_CKM00114700023_00014 is paired with replication device clariion_CKM00114700023_MZ71-0014-1775.

04/03/2013 16:59:33 MZ71 Original device clariion_CKM00114700023_00015 is paired with replication device clariion_CKM00114700023_MZ71-0015-1775.

04/03/2013 16:59:33 MZ71 Original device clariion_CKM00114700023_00018 is paired with replication device clariion_CKM00114700023_MZ71-0018-1775.

04/03/2013 16:59:33 MZ71 Original device clariion_CKM00114700023_00019 is paired with replication device clariion_CKM00114700023_MZ71-0019-1775.

04/03/2013 16:59:33 MZ71 Original device clariion_CKM00114700023_00020 is paired with replication device clariion_CKM00114700023_MZ71-0020-1775.

04/03/2013 16:59:33 MZ71 Original device clariion_CKM00114700023_00021 is paired with replication device clariion_CKM00114700023_MZ71-0021-1775.

04/03/2013 16:59:55 mz30 100510 INFORMATION: Start SQL Server Replication validation.

04/03/2013 16:59:55 mz30 100701 INFORMATION: Sufficient space for temporary VDI file on C:\tempUndoDir\

04/03/2013 16:59:55 mz30 100687 INFORMATION: Not running user scripts, skip checking user credentials

04/03/2013 16:59:55 mz30 100512 INFORMATION: Replication validation complete with no warnings

04/03/2013 16:59:55 mz30 100511 INFORMATION: End SQLServer Replication validation.

04/03/2013 16:59:55 mz30 Replication validation The volumes for this replication are all on basic disk.

04/03/2013 17:00:23 mz30 000570 ERROR: Attempt to do a consistent split of 9 LUNs on storage array CKM00114700023 which can only split 8 consistently

04/03/2013 17:00:23 mz30 026084 ERROR: Storage Services operation Validate failed with an error as follows: Storage Services has terminated with an error.

04/03/2013 17:00:22 MZ71 102206 ERROR: Validation failed.

04/03/2013 17:00:22 MZ71 102206 ERROR: Validation failed.

-----

  • etaljic81

    1000 Posts

    575

    0

    Posted March 6th, 2013 05:00

    The consistent split will split all of the LUNs you have defined in the application set at the same time. Instead, let VDI quiesce each database/logs as need be.

    The steps you provided are perfect. Use those setting and you'll be all set.

  • etaljic81

    1000 Posts

    575

    0

    Posted March 5th, 2013 07:00

    Disable consistent split and run the job again.

  • mtmmza

    6 Posts

    575

    0

    Posted March 5th, 2013 09:00

    I do that, and works. LUN's 11, 12, 13, 14, 15, 18, 19 and 20 are POOL type and 21 are RAID type. I read that maybe there is the problem. I'm reading about sesseions, but i think that RM do sessions automatically...

  • etaljic81

    1000 Posts

    575

    0

    Posted March 5th, 2013 09:00

    No need to do consistent split. RM uses VDI to quiesce the SQL application and thus you don't need to use consistent split.

    The issue is that you cannot perform a consistent split across more than 8 LUNs. You are trying to perform a consistent split on 9 LUNs and that's why the job is failing. But again, no need to use it unless you really need all 9 LUNs to be consistently split at the same exact time. Let VDI do the work and quiesce the databases as needed.

  • mtmmza

    6 Posts

    575

    0

    Posted March 5th, 2013 09:00

    8 Pool LUN's are for mdf's and RAID LUN is for ldf.

    You say that is redundant use VDI and consistent split??

    I should:

    In Advanced Replication Settings -> General Rep Options: Uncheck "Use consistent split"

    In Advanced Replication Settings -> SQL Server 2008 Instance Name:

         -Consistency method: "by instance"

         -Full, Online with advanced recovery (using VDI)

    Thanks in advance.

    Diego Santoni

  • mtmmza

    6 Posts

    575

    0

    Posted March 6th, 2013 05:00

    Thanks Ernes for your time!

    Now works perfect.

  • etaljic81

    1000 Posts

    575

    0

    Posted March 6th, 2013 05:00

    Any time, glad I can help.