1

Currently, our SharePoint farm has restricted conditions as below

  • Shared folder is not allowed
  • FTP/SFTP is not allowed
  • SQL FILESTREAM is not allowed
  • SMB protocol is not allowed
  • Using Windows Server Failover Clustering (Not SQL AlwaysOn FCI) for SharePoint database servers

I want to setup a DR farm for the existing production. First, we tried with log shipping but it turned out that log shipping does not co-exist with Failover Clustering, as stated here https://msdn.microsoft.com/en-us/library/bb500117(v=sql.105).aspx

Then, I came up with idea of using transactional database replication, but still, have considerations that couldn't find elsewhere

  • Does Transactional Replication require shared folder? (For example, to copy log file)?
  • Can Transactional Replication co-exist with Failover Clustering?

Please propose if there's any possible Disaster Recovery solution for this requirements

We don't use Failover Cluster at instance level but server level

Improve this question
1
  • 1
    I'm confused. On the page you linked, where does it say you can't use Log Shipping with a FCI? That's a very popular choice for HA/DR. – Erik Darling Jul 13 '17 at 11:16
0

Note: This part of the answer corrects the misconception about Log Shipping not being compatible with Failover Clustering.

As @sp_BlitzErik implies in the comments, Log Shipping is very much compatible with Failover Clustering. I've seen others fall prey to this implied lack of interoperability as it seems the MSDN article you link to is a direct reference of this MS Blog post from 2013. It seems even @Shanky commented about this discrepancy directly on that article as well.

Long story short, that article is incomplete. It should also reference Failover Clustering under Log Shipping along with a handful of other DR technologies.

Final note on using log shipping with Failover Clusters; the main caveat is to ensure your TLogs are being backed up to a clustered share that can failover along with the instance and other clustered services.

Answer: Now to address the rest of the answer.

You've got a few options. Really what you're able to do is often limited by your budget (e.g. licensing, hardware, etc.), so any option proposed will need to fall within your budgetary constraints. That being said, here some options I can think of, but by no means do I consider this a complete list of your available options:

  1. Log Shipping: For the sake of completeness, you CAN configure log shipping on a failover cluster. As I say above, be sure to configure the backup location as a shared folder that fails over along with the other clustered resources (e.g. SQL Server Service, SQL Agent, etc.).
  2. Multi-Subnet Failover Cluster: This is adding onto what it sounds like you already have, but you can include a different site into your failover cluster. Proper care will need to be taken so that automatic failover to your DR site is only done when necessary. Additional constraints are replicating the data to the DR site. Configuration of this approach is not simple and will depend a great deal on your Windows Version(s), SQL Server version, budget, etc.
  3. SAN-to-SAN Replication: This is generally a hardware based approach and will depend upon your SAN vendor. Most Enterprise-level SAN solutions support SAN-to-SAN replication and in conjunction with virtualization/etc. you can run across some elegant and expedient DR solutions.

Disaster Recovery is a complex topic and providing anything resembling a complete answer can't be provided on any forum answer. If you're new to this or feel overwhelmed, I heavily suggest reaching out to MS, a MS Partner, a Consulting Firm/Consultant, etc. and ask for help.

Improve this answer
5
  • So will I have to configure log shipping job on both nodes or the jobs would be created on 1 node only & which node will execute it is handled automatically ? – Brian Pham Jul 13 '17 at 12:54
  • No, this is not needed because the SQL Agent is also a clustered service, so any jobs defined on the Agent will also failover with it. You may be getting a Failover Clustered Instance confused with an Availability Group which is yet another/different option for High Availability. – John Eisbrener Jul 13 '17 at 12:58
  • And the SAN storage is where the transaction logs often stored right ? – Brian Pham Jul 13 '17 at 13:00
  • @ThắngPQ Correct tlogs, data, etc are stored on the SAN in a FCI, but based on your question, I'm a bit concerned that you may be in a bit over your head on this. I've edited my answer to include some of the available DR options, but I will heavily suggest you bring in an expert with experience building out DR solutions. – John Eisbrener Jul 13 '17 at 13:27
  • I forgot to clarify. Log Shipping will only work with SQL Sever FCI (at instance level) or with Windows Failover Clustering (at server level) also? – Brian Pham Jul 17 '17 at 5:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.