Compatibility

  • I have changed the compatibility of database accidentally and then reverted it back. Would that cause any problem? Please advise?

    Thanks

  • Admingod - Friday, January 25, 2019 10:29 AM

    I have changed the compatibility of database accidentally and then reverted it back. Would that cause any problem? Please advise?

    Thanks

    Depends on how long it was at a different compatibility level. Usually it isn't a problem if it was just a quick "oops" where you set it to another version and then switched it right back.

    Sue

  • Thanks Sue! Have you worked on Clustering setup?

  • Admingod - Friday, January 25, 2019 3:05 PM

    Thanks Sue! Have you worked on Clustering setup?

    Yes - you bored with replication errors and moving onto Clustering now? 🙂

    Sue

  • :-). I like replication. However, i need to setup clustering(Active/Passive) on 2 nodes, need help on list of things i need before setting up clustering. However, if i setup DR what would be the best possible option on cluster? Please advise?

  • Admingod - Friday, January 25, 2019 4:00 PM

    :-). I like replication. However, i need to setup clustering(Active/Passive) on 2 nodes, need help on list of things i need before setting up clustering. However, if i setup DR what would be the best possible option on cluster? Please advise?

    DR is a huge, huge topic. And there are different types of disasters. You'd want to define what disasters you need to plan for - hardware outage, site outage, natural disaster, user accidentally deleting critical data, etc. What you do is based on what the business needs and what they can afford. And all the systems end up having different requirements but you'd want to help the business define those things, understand the costs and identify any potential points of failure for the different options. Is it something that will all be done locally or will it involve different data centers, is it at the database level or server level, are there things that can be handled using NAS or SAN technologies, there are just so many of those type of questions you'd want to know. And you will likely have different plans for different types of disasters - you already take regular backups which are usually part of the plan for user errors in databases.
    So for this server, what's best depends on what's running on it and what the needs are for DR. You would want to figure out those needs before setting things up. Microsoft has a pretty good high level overview of what some of the options are with SQL Server:
    Business continuity and database recovery - SQL Server

    Mirroring is something you wouldn't want to look at since it's going away. Failover Cluster Instances are becoming more popular and something you may want to look into for your cluster. The downside is the use of shared storage but it's still a good option. 
    And don't overlook how important the backups you do are - it's one of the most important first steps. And include the availability of the backups - are they stored offsite, are they stored only on the same server as your databases (ouch), etc. Damage to a database is probably the most common disaster you will see. And you usually will see it at some point if you haven't yet.

    Sue

  • Sue,
    So the idea is to setup 2 node clustering i.e Active/Passive. Can you elaborate how does the two node cluster will fulfill the Redundancy and DR option. You mean Active node has to be on actual location and then passive node to be on DR location. So if i choose clustering option then will that be still feasible on VM servers instead of physical servers? So i need 2 servers sharing the same storage then i need two ip address for different nodes, since they would be on same cluster name right? Please let me know what else i am missing?

  • Admingod - Monday, January 28, 2019 2:34 PM

    Sue,
    So the idea is to setup 2 node clustering i.e Active/Passive. Can you elaborate how does the two node cluster will fulfill the Redundancy and DR option. You mean Active node has to be on actual location and then passive node to be on DR location. So if i choose clustering option then will that be still feasible on VM servers instead of physical servers? So i need 2 servers sharing the same storage then i need two ip address for different nodes, since they would be on same cluster name right? Please let me know what else i am missing?

    You could do it with different locations but you need to replicated the storage and how that is done and how easily depends upon the vendor as well as the existing setup you have to your DR site. This article has a pretty good diagram and explanation on how things would work:
    SQL Server 2012 Multi-Subnet Cluster Part 1

    Sue

Viewing 8 posts - 1 through 7 (of 7 total)

You must be logged in to reply to this topic. Login to reply