Inspring by a conversation seen on twitter – Travel tips from #sqlfamily
The rolling method for packing, WOW does it save space.
Eagle Creek compression cubes – https://t.co/9cjM0wc53w
“You can even send Amazon/Nordstrom orders to your ‘closet’ if you buy new clothes.”
Windows Server 2016 Technical Preview 4 Cluster Build and SQL Server CTP 3.2 – Basic Availablity GroupPosted: January 14, 2016
Windows Server 2016 Technical Preview 4 Cluster Build and SQL Server CTP 3.2 – Basic Availablity Group
Basic Availability Group Setup with a Active Directory-Detached Cluster
– iSCSI disk setup
– Firewall rule setup for default Availability Group port
– DNS Setup
– Active Directory-Detached Cluster
– Enabling SQL Authentication
– Certificate Setup including Backup/Restore with Backup Encryption
– Certificate Checks before AG setup
– Database Backup/Restore
– Creating the Basic Availability Group via the GUI
– Availability Group Checks via DMV
– Availability Failover/failback
Windows Server 2016 Technical Preview 4 Cluster Build and SQL Server Installation – standalone SQL installs with Storage Replica,floating IP and manual failoverPosted: January 12, 2016
Following on from the previious attempt at Windows Server 2016 Storage Replica with Cluster Shared Volumes which did not work completely I tried a manual approach.
I did not give up, what if we do not use Windows Clustering but do all the steps manually?
This is 2 standalone SQL Servers both installed with data on the F: drive and then Storage Replica used to keep them in sync.
– add a DNS entry and a floating IP address
– manually setup Storage Replica
– manually handle stopping/starting services, failing over/back the Storage Replica and moving the floating IP
does it work?
The answer is yes-ish if you are careful and ignore the warning about data loss!
There does not seem to be an option to failover (Set-SRPartnership) synchronously to guarantee no data loss!
Also we could do with an option to validate everything is in sync or resync even if we would have to suspend writes to allow this.
Windows Server 2016 Technical Preview 4 Cluster Build and SQL Server Installation – standalone SQL installs with Storage Replica,floating IP and manual failover
Useful tools for checking SQL Server connectivity – PortQry and PortQryUI
Powershell to manage Storage Replicas
Get-Module -ListAvailable : List available modules
Get-Module : List modules loaded
Get-Module | ft Version,Name
Version Name ------- ---- 188.8.131.52 Microsoft.PowerShell.Management 184.108.40.206 Microsoft.PowerShell.Utility 1.1 PSReadline 1.0 StorageReplica
Get-Command -Module StorageReplica | ft CommandType,Name
CommandType Name ----------- ---- Function Clear-SRMetadata Function Export-SRConfiguration Function Get-SRDelegation Function Get-SRGroup Function Get-SRPartnership Function Grant-SRAccess Function Grant-SRDelegation Function New-SRGroup Function New-SRPartnership Function Remove-SRGroup Function Remove-SRPartnership Function Revoke-SRDelegation Function Set-SRGroup Function Set-SRPartnership Function Suspend-SRGroup Function Sync-SRGroup Cmdlet Test-SRTopology
I have been playing with SQL Server Clustering (FCI) setup with a Widnows Server 2016 Preview 4 and Windows Storage Replica.
This involves local storage and Cluster Shared Volume with Replication Enabled.
This is incomplete – if Node 1 goes offline the SQL Server role fails!
Tools used when building Windows cluster with iscsi:
- Control Panel -> Network and Internet -> View network status and tasks including Change adapter settings
- Control Panel -> System -> Device Manager including View -> show hidden devices
- Disk Management
- Server Manager -> Manage -> Add roles and features
- Server Manager -> Tools -> Failover Cluster Manager
- Server Manager -> Tools -> iSCSI Initiator
- Server Manager->Tools->Routing and Remote Access
- Server Manager -> File and Storage Services on left panel -> iSCSI
- Server Manager->Tools-> Group Policy Management
- Server Manager->Tools-> Active Directory Users and Computers
Tip and undo hints when things go wrong!:
- Sysprep everything, see below, technically I read only DC needs sysprep but I was safe.
- diskpart – to remove GPT protected or recovery partitions – list disk – select disk N – list partitions – select partition N- delete partition override
- Server Manager->Tools-> Active Directory Users and Computers -> expand domain -> expand computers -> delete computer to allow retry – not only done on my play setup!
- Backup all VMs BEFORE installing Failover Clustering features
After a lot of reading I found Mark Russinovich’s answer.
So sysprep is only needed when 1 machine will reference another machine’s SID :
- The first domain controller in a domainhas a machine SID and this becomes the “Domain SID”.
- All DCs have the their machine SID matching the “Domain SID”.Machines
- All machines which are not a Domain Controller MUST not have a machine SID matching the “Domain SID”.
I sysprep’ed everything in my cluster builds so I have not tested this!